How it works

The business's phone number is connected to a platform that watches for unanswered inbound calls. When a call passes a configurable ring threshold — GoHighLevel, for example, allows 10 to 20 seconds — the platform sends a short SMS from that number to the caller's mobile. The message is typically one segment (160 characters or fewer): something along the lines of "Hi, this is [Business Name] — sorry we missed your call. How can we help?" Replies land in a shared inbox alongside other incoming texts and web chats. Most platforms let you set separate messages for business hours and after-hours calls.

What it costs

GoHighLevel includes missed-call text-back in all plans, which start at $97 USD per month; SMS usage is billed on top at approximately $0.0079 USD per message segment, per GoHighLevel's published LC Phone pricing. Most Thunder Bay businesses access GoHighLevel through a local agency or reseller, so effective pricing varies — confirm with any reseller you work with. Enzak is a standalone tool built specifically for this feature: $99 USD per month for up to 1,000 texts per month, and it works with any existing business phone number through call forwarding, with no CRM or platform migration required. Podium and Birdeye include the feature within broader communication suites but do not publish pricing publicly; confirm current rates directly with each provider.

The Canadian compliance check

Canada's Anti-Spam Legislation (CASL) covers automated commercial text messages. A single SMS sent in direct response to an inbound call is generally read as low-risk under the CASL provision for messages sent in response to customer inquiries, but that does not mean the standard requirements disappear: every commercial text must identify the sender by name and include an unsubscribe mechanism — "Reply STOP" is the standard, with "ARRET" for French — and opt-out requests must be honoured within 10 business days. Enrolling the caller into ongoing marketing sequences is a separate step that requires explicit consent. If you are unsure how CASL applies to your specific setup, confirm with a legal adviser before going live.

What missed-call text-back does not do: it does not answer the phone, screen callers, or resolve what someone called about. It is a holding action, not a resolution. A caller who dialled the wrong number still gets the text. Keep the message short, specific, and realistic — "a plumber will call you back within the hour" works better than a generic form letter. The automation buys time; following through is still the job.

Which Northwestern Ontario businesses benefit most

Any service business where calls arrive during busy or unmanned windows sees the clearest return. Trades — plumbers, electricians, HVAC — field calls while technicians are on job sites. Health clinics miss calls when front-desk staff are with patients. Restaurants and lodges receive reservation enquiries during peak service hours. In each case, the text-back does not replace the follow-up call — it tells the caller the business received their call and creates a channel for them to wait rather than immediately trying the next number on their list.

Frequently asked questions

• Does the caller need to install anything to receive the text? No. The SMS goes to their existing mobile number — no app, no account, nothing required on their end. The only constraint is that they must have called from a mobile number capable of receiving SMS; calls from landlines will not receive the text.
• Can I use my existing business number, or do I need a new one? Most platforms let you send from your existing number. Enzak, for example, works with any landline, VoIP, or mobile number through call forwarding. Some platforms may require you to port your number or use a provisioned one — confirm before signing up.
• What should the automated message actually say? Keep it short and honest: your business name, acknowledgment of the missed call, and a realistic expectation for when someone will follow up. Avoid generic templates. A message that sounds like it came from a person works better than one that reads like a form letter.

Meta breaks ground on Canada's first $13 billion AI data centre — in Alberta, not Ontario

Meta broke ground in Sturgeon County, Alberta on July 8 on its first Canadian data centre — a one-gigawatt facility representing more than $13 billion CAD, and Meta's largest outside the United States. With 3,000 workers expected on site at peak construction and more than 300 permanent operational jobs, it is the largest AI infrastructure commitment in Canadian history. Alberta's pitch was affordable electricity and cold temperatures that reduce cooling costs — the same two advantages Northwestern Ontario holds. Ontario and the federal government have both signalled interest in sovereign AI data centre development; a federal solicitation for 100-megawatt-plus compute facilities earlier this year drew national attention before its window closed. No NWO announcement has been made, but local commentary noted that Northwestern Ontario's hydroelectric power and cold climate position it to compete for exactly this kind of investment.

Gordie Howe Bridge gets its opening date: July 27

Canada's federal housing and infrastructure department and Michigan Governor Gretchen Whitmer jointly announced on July 10 that the $4.4-billion Gordie Howe International Bridge between Windsor and Detroit will open July 27, 2026, connecting Highway 401 in Ontario to Interstate 75 in Michigan. The crossing was originally set to open in June 2026 but was delayed to resolve outstanding issues. The bridge adds a second major span to what officials described as one of the busiest transportation corridors in North America. For Northwestern Ontario manufacturers and resource exporters whose goods move to US markets via the 401, a second crossing means more capacity and reduced congestion on the main land route between Ontario and the US Midwest.

Ontario rewards Thunder Bay with a third housing payment — $709,280

Ontario presented Thunder Bay with a $709,280 cheque on July 10 under the provincial Building Faster Fund — the third and final round of the program. Thunder Bay earned the payment by meeting its provincially designated housing target for a third consecutive year; the fund rewards municipalities that reach at least 80 percent of their assigned annual target. The Building Faster Fund is being sunsetted after this round. The city plans to use the funds for infrastructure that supports its goal of building 2,100 new housing units by 2031.

SpaceXAI releases Grok 4.5, a frontier coding model priced at $2 per million tokens

SpaceXAI released Grok 4.5 on July 8, a new large model built for coding and agentic tasks. Elon Musk described it at launch as "an Opus-class model, but faster, more token-efficient and lower cost." It is priced at $2 per million input tokens and $6 per million output tokens, available to Canadian users now; EU access is expected in mid-July. The release continues a pattern visible across the past several weeks: capable frontier-class models arriving at or below $3 per million input tokens. That compression makes AI workflows meaningfully cheaper to run on a working small-business budget, and changes the economics of automations that would have been too expensive to bother with a year ago.

Thunder Bay's planned $80 million waterfront science centre is delayed

Science North announced this week that construction of its planned $80-million science centre on Thunder Bay's downtown waterfront will take longer than originally anticipated, and has not set a revised timeline. The project, which carries a previously committed $20-million federal contribution and was expected to begin construction in spring 2027, is planned for the waterfront near the downtown art gallery development that is also on hold. Science North said its commitment to the project and to Northwestern Ontario has not changed.

What Alberta actually did

The scan covered systems across all 27 of Alberta's provincial ministries; the Ministry of Technology and Innovation says it runs more than 1,280 applications and 3,400 collections of code in total. About 50 Claude Code agents worked autonomously and in parallel — first flagging known vulnerability patterns with a rules engine, then reviewing each flag and citing the exact file and line so a human developer could verify it. Where the scan found real problems, Claude Code often went further: generating fixes, writing tests that didn't exist, and in some cases rebuilding outdated systems in more current languages. Alberta's own estimate for modernizing that technology the traditional way — not just the initial scan, the full fix — is roughly $2 billion and more than a century of work.

Then it gave the method away

This is the part that matters most for a smaller jurisdiction. Alberta didn't just do the work quietly and keep the advantage. It published 21 technical white papers — the Velocity White Papers, at thevelocitywhitepapers.com — free, open-source, with step-by-step instructions, specifically so other governments facing the same aging-systems problem don't have to redo the R&D from scratch. Alberta's Minister of Technology and Innovation, Nate Glubish, put it plainly: "Albertans trust their government with some of the most sensitive information in their lives, and it is our responsibility to protect it. By using AI to find and fix vulnerabilities across our systems, we accomplished in hours what would have taken a traditional approach years to complete. This is what responsible government looks like in the AI era, and the best is still ahead of us." The province is also planning to use AI agents to consolidate 165 separate legacy systems into 16 new applications it will own outright, rather than continuing to license and patch the old ones.

Why this is a Northwestern Ontario story

Swap "Alberta ministry" for a City of Thunder Bay department, a district health authority, a school board, or a college, and the underlying problem is identical, only smaller in scale: systems built up over decades, patched by whoever was available at the time, with no complete picture of what is actually running or how exposed it is. That kind of review has always been theoretically possible and practically impossible — no NWO municipality or public institution has the budget to hire a team for a multi-year manual code audit. Alberta just demonstrated, at a scale no one can dismiss as a toy example, that the practical-impossibility half of that equation has moved substantially. The playbook for how it did that is now public and free to read.

What this doesn't mean: a push-button fix. Alberta ran this with an in-house technical team inside a ministry built for the work, using a purpose-built coding tool, and every flagged fix still needed a human who understood the codebase to verify it before it shipped. The 466-million-line figure and the cost/time estimates are Alberta's own reporting, not an independently audited number — treat them as the province's claim, not a settled fact. A Northwestern Ontario institution attempting anything similar needs real in-house technical capacity or a partner who can run and interpret the results responsibly. A free playbook does not make the labour to use it free.

A practical first step

The realistic move for an NWO institution is not to attempt Alberta's scale — it's to read what Alberta published before committing to a vendor or a system rebuild sight-unseen. The white papers are public at Alberta's own site, and using them to scope one system or one repository as a pilot is a cheaper, lower-risk starting point than commissioning a study from zero. From there, loop in whoever actually maintains your systems — an internal IT lead, if you have one, or a local implementation partner if you don't — and treat the pilot's findings, not a vendor's pitch, as the basis for what to fund next. If funding is the gap, confirm current eligibility directly with the relevant program before assuming it applies to a systems-review project; see thunderbayai.com's funding page for what is currently active in the Northwest.

The three models, and which one a small business actually needs

GPT-5.6 is not one model — it is a tiered family, and the tier you choose is mostly a cost decision. Most local business tasks do not need the flagship.

• Sol — the flagship, priced at $5 per million input tokens and $30 per million output. Built for frontier reasoning and long-horizon agentic work: tasks with many steps, tool use, and planning. OpenAI adds a new "max" reasoning effort and an "ultra" mode for the hardest problems. For most small businesses this is more model than the job requires.
• Terra — the balanced middle tier at $2.50 input / $15 output. OpenAI positions it as competitive with the previous generation's everyday model at a lower cost. This is the sensible default for drafting, summarizing, answering customer questions, and general assistant work.
• Luna — the fastest and cheapest at $1 input / $6 output. Built for high-volume, straightforward work: classifying incoming messages, tagging leads, pulling fields out of documents, short replies. At this price, automations that were not worth building a year ago start to pencil out.

Why the price drop is the real story for a local operator

The number that matters for a Thunder Bay shop, clinic, or trades business is the cost per token, because that is what you actually pay to run an AI feature day after day. OpenAI CEO Sam Altman said Sol is 54 percent more token-efficient on coding tasks than its predecessor — fewer tokens for the same result means lower running cost. Combine that with Luna at a dollar per million input tokens, and the arithmetic changes: an assistant that reads and triages every inbound email, or a process that extracts details from hundreds of invoices a month, moves from "too expensive to bother" into "cheap enough to just turn on." The right move is rarely to reach for the most powerful model. It is to run the smallest, cheapest model that clears the quality bar for the task, and to reserve the flagship for the genuinely hard, multi-step jobs.

The government held the release — and that should sound familiar here

GPT-5.6 did not ship on OpenAI's original timeline. The Trump administration sought to restrict its rollout over cybersecurity concerns, tied to Sol's strength in dual-use domains like biology, chemistry, and offensive cyber work. Under a June 2026 federal executive order, major AI developers were asked to voluntarily submit their leading models to government regulators for safety evaluation before public release — a policy shaped in part by earlier concern over Anthropic's Mythos model. OpenAI ran that review, then announced on July 8 that it was expanding preview access globally, while stating it did not believe government evaluation should become the long-term default. This is the same pattern that took Anthropic's Fable 5 offline for foreign users for 19 days in June, which we covered when access was restored. The lesson for a Northwestern Ontario business is not to follow US politics — it is that a decision made in Washington can reach into a tool you depend on in Thunder Bay, on short notice.

A caveat worth holding onto: the capability and efficiency figures here are OpenAI's own claims at launch; independent benchmarks take time to catch up, so treat them as vendor numbers until third parties verify them. Pricing, model names, and regional availability can change — confirm the current figures on OpenAI's own pricing page before you budget around them. And the practical takeaway from both the GPT-5.6 review and the Fable 5 shutdown is the same: do not build a business-critical workflow on a single model or provider you cannot swap out if access changes.

What else OpenAI shipped the same week

Two other releases landed alongside the models. OpenAI launched ChatGPT Work, a workplace assistant for teams that runs on desktop, web, and mobile and is aimed at everyday clerical tasks. And on July 8 it released new voice models, GPT-Live-1 and GPT-Live-1 mini, built for more natural spoken conversation and better turn-taking — relevant if you are considering an AI phone assistant or voice-based customer service, though as with any new model, test it against your own calls before trusting it with customers.

What a Northwestern Ontario business should actually do about it

• Do not rush to the flagship. Start with Terra or Luna and only move up if the cheaper tier misses on quality for your specific task.
• Revisit automations you shelved as too costly. High-volume, low-complexity work — message triage, tagging, document extraction — is now dramatically cheaper to run.
• Keep provider-portability. Design your workflow so you can swap the underlying model, so a price change, an outage, or a government restriction does not take your operation down with it.
• Do not rebuild around a preview. Access is still expanding; confirm availability and pricing directly with OpenAI before committing a critical process to a specific model.

Frequently asked questions

• Does my business need Sol, the top model? Probably not. Sol is built for hard, multi-step reasoning and agentic tasks. Most local use cases — drafting, summarizing, answering customer questions, classifying messages — run well on Terra or Luna at a fraction of the cost. Match the model tier to the difficulty of the task, not to the marketing.
• Is GPT-5.6 available in Canada? OpenAI said at launch on July 8 that it was expanding preview access globally, and the models went to public release on July 9. Access and pricing can still be rolling out and can change — confirm current availability for your region and plan directly on OpenAI's site before you build around it.
• Could a US restriction cut off my access the way Fable 5 was pulled? It is a real, if uncommon, risk — the GPT-5.6 review and the June Fable 5 shutdown both show that access to a frontier model can be gated or paused by government action. The practical protection is not prediction; it is portability. Keep your workflow able to switch models or providers so a single disruption does not stop your business.

Start with the task, not the model

The right tier is a property of the job, not a badge of quality. A flagship model is built to reason through many steps, plan, use tools, and hold a long context — real capability you pay a premium for. But most business work is not that. Classifying an email, pulling a total off an invoice, or drafting a reply to a common question is straightforward, and a small, cheap model handles it as well as an expensive one. Match the model to the difficulty of the task, and you stop paying flagship prices for work a budget model nails.

A tier for each common job

Using the current GPT-5.6 family as the example, here is how the tiers map to the work a Northwestern Ontario business actually does. The same shape — a cheap small tier, a balanced middle, and a premium flagship — holds across most providers.

• Cheapest / fastest tier (e.g. Luna, $1 / $6 per million tokens): high-volume, low-judgment work — sorting and tagging incoming messages, routing leads, extracting fields from documents, short canned replies, first-pass summaries. If you are running the task hundreds or thousands of times a month, this is where the savings live.
• Balanced middle tier (e.g. Terra, $2.50 / $15): the everyday default — customer chat, drafting emails, quotes, and product descriptions, summarizing longer documents, answering FAQs. Good enough for almost all customer-facing and back-office writing.
• Flagship tier (e.g. Sol, $5 / $30): complex, multi-step reasoning and agentic tasks — analysis that chains many steps, code, planning across tools. Reach for it only when the cheaper tier visibly falls short on a specific job, not as a starting point.

How to actually decide: a three-step test

You do not need benchmarks to choose. You need a sample of your own work and ten minutes.

• Run the task on the cheapest tier first. Start at the bottom, not the top.
• Check a real sample of the outputs against your own standard — not a demo, your actual emails, invoices, or customer questions.
• Move up a tier only where the cheaper model visibly fails. If you cannot tell the difference between the cheap and the expensive output, use the cheap one. The price gap is real; the quality gap on a simple task usually is not.

Two caveats. First, quality varies by task, not just by tier — a small model can match a flagship on drafting and still fall short on multi-step analysis, so test on your own work rather than trusting a general ranking. Second, prices, tier names, and availability change fast: the figures here are the GPT-5.6 line as of July 2026, and you should confirm current pricing on the provider's own site before you budget. The durable point is the method, not the numbers: cheapest tier that clears the bar, and keep your workflow able to switch models so a price change or an outage does not lock you in.

Why this matters more for a small Northwestern Ontario business

A large company can absorb an oversized AI bill without noticing. A Thunder Bay shop, clinic, or trades business runs on tighter margins, and the cost that matters is not the one-time setup but what a feature costs to run every day. Choosing the right tier is the difference between an automation that quietly pays for itself and one that eats the savings it was supposed to create. It also changes what is worth building at all: at a dollar per million tokens, a process that reads and sorts every inbound email, or extracts details from hundreds of documents a month, moves from too expensive to bother into cheap enough to just turn on. Picking the tier well is what makes the cheap end of that range usable.

Frequently asked questions

• Isn't the newest, biggest model always the best choice? No. Bigger models cost more and often run slower, and for most business tasks a mid or small tier is hard to tell apart from the flagship. The biggest model is the right default only for genuinely hard, multi-step reasoning — not for everyday drafting, chat, or classification.
• How do I know if a cheaper model is good enough? Test it on a real sample of your own work and judge the outputs against your own standard. If you cannot reliably tell the cheap tier's output from the expensive one on your actual tasks, the cheap tier is good enough — use it.
• Does this apply to models other than OpenAI's? Yes. The tiering pattern — a small, cheap model, a balanced middle, and a premium flagship — is industry-wide across the major providers. The pricing here is one current example, but the decision method is the same regardless of which provider you use, and keeping your setup able to switch providers is part of using it well.

What the program pays — and to whom

The program distinguishes sharply between for-profit businesses and public or community-sector organizations. For-profit businesses receive coverage of up to 50 percent of the intern's salary plus MERCs, to the program maximum of $35,000 per position per year. Municipalities, Indigenous communities, non-profit organizations, colleges, universities, and research institutions receive coverage of up to 90 percent of salary plus MERCs, to the same $35,000 ceiling. In both cases, the program funds salary costs only — discretionary benefits, training costs, equipment, and travel are not eligible expenses. NOHFC disburses funding in installments: one payment after the six-month mark and a final payment upon completion of the internship, both based on actual costs incurred.

Eligible organizations: who can apply

The program is open to for-profit businesses, municipalities, Indigenous communities, non-profit organizations, colleges, universities, and research institutions operating in Northern Ontario. To be eligible, the organization must have been in operation for at least one year and maintain a minimum of one full-time employee. Recruitment must be fair and transparent — positions must be advertised through public channels, not filled privately. Hiring immediate family members of owners or directors is generally not permitted, with limited exceptions in small and remote communities. Contact NOHFC directly to determine whether your organization type and hiring situation fall within the program's scope.

Roles that qualify — and roles that do not

The program is built for positions that require meaningful skill development over an extended period. Technology roles — software development, data analysis, IT administration, cybersecurity, digital marketing — are the type of position the program is designed to support, along with skilled trades, professional services, and research roles. A business hiring its first dedicated AI or software specialist, a municipality bringing on a GIS or data analyst, or a non-profit adding a digital communications position are the kind of use cases that fit.

• Positions requiring minimal training for fewer than six months are not eligible.
• General manual labour roles do not qualify.
• Clerical and administrative positions are excluded.
• Retail and sales positions are excluded.
• Positions that would displace or replace an existing employee are not eligible.

Intern requirements

The person filling the funded position must meet all of the following criteria, as stated by NOHFC:

• At least 18 years of age.
• Reside in Canada and be legally entitled to work in Canada.
• Be a new entrant to the workforce, a career-changer, or an unemployed or underemployed person entering a new field — the program targets individuals building new skills, not those already established in the role.
• Must not have previously participated in an NOHFC-funded internship.
• Must work a minimum of 35 hours per week for at least 52 weeks.
• Must receive direct on-site supervision from the host organization.

Maximum of two funded positions per organization

Standard organizations — businesses, municipalities, Indigenous communities, and non-profits — can hold a maximum of two NOHFC Workforce Development positions at the same time. Colleges, universities, and research institutions are eligible for up to five research-specific positions in addition to the standard two, for a combined maximum of seven. Organizations that have previously held funded positions can apply for additional rounds, provided each new intern meets all eligibility requirements independently.

How to apply

Applications are submitted through NOHFC and reviewed in a multi-step process: an initial submission evaluated against program guidelines, a detailed review phase, and a final decision by the NOHFC Board of Directors. Intake is continuous — there is no fixed deadline — but funding is limited and meeting the eligibility criteria does not guarantee an approved position. Contact NOHFC at 705-945-6700 or through the inquiry portal at myportal.nohfc.ca to confirm your organization's eligibility, identify whether the role you want to fill qualifies, and understand what the application package requires. Reaching out before posting the position is the practical way to avoid building a hiring timeline around a program that may not cover your specific situation.

Continuous intake does not mean unlimited funding: the program is competitive and budget-constrained. The program covers salary costs only — benefits, equipment, training, and travel are not eligible. Funded interns must not have previously held an NOHFC-funded position. The 50 percent coverage rate for businesses applies to salary plus MERCs, to a maximum of $35,000 per position per year. Confirm your organization's eligibility and the current availability of program funding directly with NOHFC before building a hiring plan around this program.

Frequently asked questions

• Can the program fund an AI or software development role? Technology positions requiring sustained skill development over the internship period are the type of role the program is built for. Whether a specific position — say, a machine learning developer, a data analyst, or a software engineer — qualifies depends on how it is structured and scoped. Confirm with NOHFC that the role description and supervision arrangement meet the program's requirements before proceeding.
• Can a business combine this with the SR&ED tax credit? The NOHFC Workforce Development Program covers salary costs for the intern position; SR&ED covers qualifying R&D expenditures, including eligible salary costs on qualifying scientific or technical work. If the intern is performing qualifying R&D, there could be interaction between the two programs. A qualified SR&ED preparer can assess whether the intern's work meets the SR&ED definition of eligible experimentation and whether stacking is appropriate for your specific situation.
• Does the intern need to be a new graduate? No. The program targets new entrants to the workforce, career-changers, and unemployed or underemployed individuals entering a new field. A mid-career professional pivoting into a technology role, or an experienced trades worker transitioning into a new sector, could qualify — provided they have not previously held an NOHFC-funded internship and meet all other program criteria. Confirm the specific candidate's eligibility with NOHFC.

Two funding tiers: research-only versus R&D plus commercialization

The program distinguishes between two types of projects, each with its own funding ceiling. Applied Research and Development projects — work focused on developing a new technology, product, or process — are eligible for up to $500,000. Projects that extend into demonstration and commercialization, or that combine all three phases (R&D, demonstration, and commercialization), are eligible for up to $2 million. In both cases, NOHFC covers up to 50 percent of eligible costs. The difference matters at the planning stage: if your project will move from research into bringing a product to market, structuring the application to cover the full scope can access the higher ceiling. Confirm with NOHFC how your project scope maps to their definitions before finalizing an application.

What the Innovation program funds

NOHFC lists the following as eligible cost categories for Invest North Innovation:

• Direct internal technical labour: specialized work performed in Northern Ontario by employees with relevant technical expertise.
• Prototyping and design services: costs of building and testing prototypes of the new technology or product.
• Product testing and certification: third-party testing, regulatory approval, and certification processes.
• Intellectual property protection: patent applications, trademark registrations, and related legal costs to protect the innovation.
• Building and renovation costs: capital expenses for facilities directly supporting the commercialization of the technology.
• Material costs: physical materials used in the R&D or demonstration work.
• Marketing — capped: in-person or virtual trade show support and marketing material design, limited to 20 percent of total eligible project costs.

What the program does not fund

Several cost categories are explicitly ineligible. Projects aimed solely at improving a business's own internal operations — rather than developing a new product or technology for the market — fall outside the program. Ineligible costs include land acquisition, working capital, administrative and overhead expenses, the labour of owners holding 10 percent or more ownership stake, in-kind contributions, business plan development, travel, standard office equipment, and vehicles. The program is built for technology development with a commercial outcome, not internal efficiency upgrades.

The 75 percent government funding cap

While NOHFC covers up to 50 percent of eligible costs, the program imposes a hard ceiling: total government funding from all sources cannot exceed 75 percent of total eligible project costs. This means the applicant must contribute at least 25 percent of total eligible costs from non-government sources. If the project is also drawing on federal programs — such as NRC IRAP, FedNor RAII, or the SR&ED tax credit — the combined public contributions across all programs must stay at or below 75 percent. Mapping your funding stack against this cap is a practical step before finalizing the application, since a project drawing on multiple programs can run into the ceiling without careful planning.

Academic and research institution partnerships

Invest North Innovation explicitly permits partnerships between private-sector businesses and public-sector academic or research institutions — universities, colleges, and research centres. Lakehead University and Confederation College, both based in Thunder Bay, are examples of institutions where these partnerships are possible. In a partnership structure, the private-sector business must be the lead applicant and the direct recipient of the NOHFC funding; the academic partner is a collaborator, not a co-applicant. Partnerships can strengthen an application by demonstrating technical capacity and research credibility — but the private entity remains responsible for the project and for meeting the program conditions.

How to apply

Applications go through NOHFC and proceed in three stages: initial submission reviewed against program guidelines, a detailed evaluation phase in which NOHFC may request additional information, and a final decision by the NOHFC Board of Directors. Intake is continuous, but funding availability is limited and the board exercises discretion — qualifying for the eligibility criteria does not guarantee funding, and not every qualifying project is approved. Contact NOHFC at 705-945-6700 or through the inquiry portal at myportal.nohfc.ca before building a full application to confirm your eligibility, your project scope, and what the submission package should include. Engaging NOHFC early is the practical way to identify whether a specific cost category or project structure falls inside the program before investing time in the full submission.

Invest North Innovation is a competitive program: meeting the eligibility criteria does not guarantee funding. The conditional contribution is not an unconditional grant — conditions around keeping the project and funded assets in Northern Ontario, meeting employment and reporting requirements, and complying with your funding agreement all apply. Total government funding from all sources cannot exceed 75 percent of eligible project costs. Projects aimed solely at internal efficiency are not eligible. Confirm your specific eligibility and the current intake status with NOHFC before applying.

Frequently asked questions

• Can a software company apply? Private-sector businesses developing new software products or technology solutions are the type of applicant the program is designed for, provided the work constitutes genuine R&D or commercialization rather than internal operational tooling. Confirm with NOHFC whether your specific project meets the program's definition.
• Can the grant and SR&ED be stacked? The SR&ED tax credit and Invest North Innovation address different cost bases and operate under different rules, so stacking is sometimes possible — but total government funding from all sources cannot exceed 75 percent of eligible project costs. A qualified SR&ED practitioner can help map the interaction.
• Does the business need to be based in Northern Ontario? Yes. Work must be performed in Northern Ontario, and NOHFC requires applicants to confirm they fall within the program's eligible geographic boundaries. Contact NOHFC to verify your location qualifies.

Fable 5 is back — the 19-day AI model ban is over for Canadian users

On July 1, Anthropic restored global access to Claude Fable 5 and Mythos 5 after the US Commerce Department lifted the export controls imposed on June 12. The original order cut off both models for all foreign nationals worldwide within days of launch, triggered by a jailbreak technique that Amazon researchers found allowed the model to identify software vulnerabilities. Anthropic's fix is a new safety classifier that blocks the specific jailbreak in over 99 percent of cases; users who attempt the blocked technique are redirected to Opus 4.8. Access is now restored through Claude.ai, Claude Platform, Claude Code, and Claude Cowork. For Canadian users who were mid-pilot when access cut out on June 12, the practical immediate effect is that Fable 5 is available again. The structural condition that enabled the shutdown is unchanged: as part of restoring access, Anthropic committed to providing US government partners early access to frontier models before public release and to co-developing jailbreak risk scoring with major cloud partners. The 19-day block resolved quickly — it is worth noting that the framework enabling a repeat is now more formalized, not less.

FedNor opens a $39M community infrastructure stream — Northern Ontario municipalities and Indigenous organizations can apply now

On June 30, Minister Patty Hajdu announced at the Canada Games Complex in Thunder Bay that FedNor is delivering $39 million to Northern Ontario through the Local Impact stream of the federal Build Communities Strong Fund. The funding is for municipalities and Indigenous organizations to build, repair, and upgrade the community facilities people rely on daily — centres, parks, cultural spaces, and local infrastructure. The $39 million rolls out over four years. The press release did not list a specific intake deadline; applications go directly through FedNor. This stream targets civic and community infrastructure rather than commercial ventures, and is separate from FedNor's business innovation and economic development programs.

74 active wildfires across the Northwest as of July 4 — conditions deteriorating

As of 6:24 PM CDT on July 4, Ontario's Northwest Fire Region had 74 active wildfires, up sharply from 13 on June 25. Seventeen of those fires are uncontrolled. Three complexes are driving the largest resource demands: Fort Frances 14, an uncontrolled fire of 1,480 hectares near Byers Lake approximately 35 kilometres southwest of Upsala, with 14 crews and 5 helicopters assigned; the Kasabonika Lake First Nation Complex, multiple fires totalling more than 3,900 hectares combined with 12 crews, 8 additional personnel, 3 helicopters, and 2 fixed-wing aircraft deployed; and Sioux Lookout 21, 357 hectares and uncontrolled near Wunnumin Lake, with 6 crews and 3 helicopters. Three new smaller fires emerged near Lake Nipigon on July 4. A drying trend is expected to increase fire behaviour as fuel moisture decreases. Forestry, construction, and any outdoor operations across the Northwest should monitor current fire status and road closures through the provincial fire information portal.

July 31 deadline for Northern Ontario municipalities: NORDS Fund closes in 25 days

The Ontario government's Northern Ontario Resource Development Support Fund closes for applications on July 31, 2026 at 5 PM EDT. The program provides $15 million annually over five years to help Northern Ontario municipalities offset infrastructure costs tied to resource-sector activity — roads, bridges, and capital assets that carry mining and forestry traffic. All 144 Northern Ontario municipalities are eligible, including those in the Thunder Bay, Kenora, Rainy River, Algoma, Cochrane, Sudbury, and adjacent districts. Individual municipalities can receive up to $400,000 per year, and the program can fund up to 100 percent of eligible capital costs. Only capital expenses incurred on or after April 1, 2026 qualify. Municipalities with resource-sector infrastructure pressures and a relevant capital project should confirm eligibility with the Ministry and submit before the deadline.

Canada tables its first AI-era privacy law — automated decision-making transparency is now in scope

On June 15, Minister of Artificial Intelligence and Digital Innovation Evan Solomon tabled Bill C-36, the Protecting Privacy and Consumer Data Act — the most significant overhaul of Canada's federal private-sector privacy framework in 25 years. The bill is before Parliament and has not yet passed into law. Key provisions include: organizations must explain automated decision-making in plain language; individuals gain the right to request deletion of their personal data; data portability is required; surveillance pricing is prohibited; and children's data receives enhanced protections. A new Digital Safety and Data Protection Commission would enforce the legislation, with fines of up to $10 million or 3 percent of global revenue, rising to $25 million or 5 percent of global revenue for serious violations. For any NWO business using AI tools to screen applicants, set prices, or make decisions affecting individual customers, the transparency and consent requirements in Bill C-36 mark the direction of travel — building those disclosure practices ahead of royal assent puts a business well ahead of the compliance curve.

Three funding structures: which one fits your expansion

• Grant only: NOHFC contributes up to 20 percent of eligible costs as a conditional contribution, to a maximum of $400,000. This is the non-repayable component — you are not required to pay it back provided you meet the program conditions.
• Grant plus loan (hybrid): The conditional contribution covers up to 20 percent of eligible costs (maximum $400,000), and NOHFC adds a term loan covering up to 30 percent of eligible costs (maximum $600,000). Combined maximum across both components is $1 million.
• Loan only: A term loan covering up to 50 percent of eligible costs, to a maximum of $1,000,000, with no grant component. Repayable on agreed terms.

Which structure fits your project depends on the size of the expansion, your cash position, and how much leverage you want to carry. A smaller capital project with strong owner equity might fit the grant-only track. A larger facility build or significant equipment purchase might warrant the hybrid or loan-only option to reach the funding threshold the project needs. NOHFC reviews applications competitively within each quarterly intake — qualifying for the eligibility criteria does not guarantee funding, and not every qualifying project is approved.

What Invest North Grow funds

The program covers capital costs tied to expanding existing Northern Ontario operations. Eligible cost categories listed by NOHFC include: building construction and facility upgrades; new or used equipment purchases; IT and communications investments; land development directly supporting the expansion; product marketing, capped at 20 percent of project costs or $75,000 (whichever is lower); and third-party training, limited to 20 percent of project costs.

What the program does not cover: working capital, ongoing operating expenses, land purchases, and land development intended for resale or lease. A project must also demonstrate that it will create or retain full-time jobs in Northern Ontario — that employment outcome is a core program requirement evaluated as part of the application, not a secondary consideration.

Who qualifies — and who does not

Invest North Grow targets established private-sector businesses with an expansion project in Northern Ontario. The owner or primary applicant must work full-time in the business. This is an expansion program, not a startup program — if your business has been operating less than six months, the Invest North Launch stream is the relevant one.

• Retail and consumer-service businesses are not eligible.
• Accommodation businesses, food service operations, and beverage-alcohol manufacturers (distilleries, breweries, cideries, wineries) are excluded.
• Land purchase costs are not eligible project costs.
• Projects that do not create or retain full-time jobs do not meet the core program requirement.
• Total government funding from all sources — federal, provincial, and municipal — cannot exceed 50 percent of eligible capital costs.

Some business types that fall outside the standard eligibility may be considered case-by-case: professional services filling a genuine gap in a small community where the service is not otherwise available full-time locally, and remote tourism operations such as outfitters and vacation resorts serving longer-stay visitors. Confirm your specific situation with NOHFC directly before assuming you fall inside or outside eligibility.

Quarterly intake: four deadlines per year

Invest North Grow runs on a competitive quarterly schedule rather than a truly open rolling intake. The four annual submission deadlines are April 30, July 31, October 31, and January 31. Missing a deadline means waiting roughly 90 days for the next round. Applications submitted before each deadline compete against each other within that round; NOHFC selects projects based on the strength of the application and available budget. The next deadline following the date of this post is July 31, 2026. Verify the current intake status and any changes to submission requirements with NOHFC before preparing an application package.

Is the grant portion repayable?

The conditional contribution component is not a loan with a fixed repayment schedule, but it is not an unconditional cash transfer either. NOHFC describes it as a conditional contribution: the funds do not need to be repaid provided you fulfil the program conditions — typically maintaining operations in Northern Ontario, keeping funded assets in the region for the required period, retaining the jobs created or preserved, and complying with the terms in your individual funding agreement. The loan component of the hybrid or loan-only options is repayable on terms set in the agreement.

How to apply

Applications are submitted through NOHFC and reviewed in a multi-step process: an initial submission reviewed against program guidelines, a detailed evaluation phase where NOHFC may request additional information, and a final decision by the NOHFC Board of Directors. Contact NOHFC at 705-945-6700 or through the inquiry portal at myportal.nohfc.ca before the deadline to confirm your eligibility and get guidance on what the application package should include. Discussing your project with NOHFC staff in advance is the practical way to verify that your business type, project scope, and intended cost categories are eligible before investing time in a full submission.

Invest North Grow is a competitive program: meeting the eligibility criteria does not guarantee funding. The grant component is a conditional contribution — conditions include retaining jobs, keeping funded assets in Northern Ontario, and complying with your funding agreement. Total government funding from all sources cannot exceed 50 percent of eligible capital costs. Retail, accommodation, food service, and wholesale businesses are not eligible. Confirm your specific eligibility and the current intake status with NOHFC before applying.

What the Launch program funds

Launch covers capital costs directly tied to starting the business. NOHFC lists the following as eligible cost categories: new or used equipment purchases; building construction and land development; leasehold improvements and leased space upgrades; IT and communications investments; marketing for new products or the new business itself, capped at 20 percent of project costs or $75,000, whichever is lower; and third-party training, limited to 20 percent of project costs. Ineligible costs include operating expenses, working capital, inventory, vehicle and boat purchases, land purchase or lease, business acquisitions, R&D, residential property improvements, mineral exploration, and electricity generation projects.

Who qualifies — and who does not

The program targets new private-sector businesses starting operations in Northern Ontario. The owner must work full-time in the business, and the business must have been operating six months or less at the date of application. Professional services businesses filling a genuine gap in a small community — where that service is not currently available full-time locally — may be considered. Tourism businesses serving longer-stay visitors, specifically remote outfitters and vacation resorts, are noted as potentially eligible by NOHFC. General accommodation and food service businesses are not eligible.

• Retail and consumer service businesses are not eligible.
• Accommodation businesses, food service operations, and wholesale businesses are excluded.
• Hotels and motels — businesses primarily providing overnight lodging — are explicitly excluded.
• Beverage-alcohol manufacturers (distilleries, breweries, wineries, cideries) are excluded.
• Businesses that have been operating more than six months cannot apply under this program.

The funding structure: 50% coverage, 15% minimum cash

NOHFC contributes up to 50 percent of total eligible project costs, to a maximum of $200,000. The applicant must contribute at least 15 percent of total eligible costs from their own resources. Combined federal and provincial government funding from all sources cannot exceed 50 percent of eligible project costs — so if you are drawing on another government program alongside Launch, both contributions together must stay at or below that cap. If funded assets are sold after project completion, NOHFC may require reimbursement. Assets cannot be moved outside Northern Ontario within three years of project completion without NOHFC consent.

Is an IT or technology investment an eligible cost?

Yes. IT and communications investments are explicitly listed as an eligible cost category by NOHFC. For a new business in Northwestern Ontario setting up its digital infrastructure from day one — hardware, business software, networking, or communications systems — Launch can contribute to that capital cost. R&D projects and business acquisitions are specifically excluded, so the program applies to deploying and setting up technology rather than developing new products. Confirm the specific costs you plan to claim with NOHFC before applying.

Is Invest North Launch a grant or a loan?

Neither, exactly. NOHFC describes it as a conditional contribution — not a repayable loan with a repayment schedule, but not an unconditional grant either. The funding does not need to be repaid as long as you meet the program conditions: maintaining full-time operations, keeping funded assets in Northern Ontario for three years, and complying with the terms in the funding agreement. If you sell funded assets after the project is complete, NOHFC may require reimbursement. The conditions and their consequences are set out in each individual funding agreement.

How to apply — and when

Applications go through NOHFC and proceed in three phases: initial submission reviewed against program guidelines, an evaluation phase with additional information gathering, and a final decision by the NOHFC Board of Directors. Intake is continuous, but funding availability is limited and the board exercises discretion — not every qualifying project receives support. The practical implication of the six-month eligibility window is that the ideal time to engage NOHFC is during the planning phase, before or very shortly after launch. Contact NOHFC at 705-945-6700 or through the inquiry form at myportal.nohfc.ca to start the conversation.

Launch is a conditional contribution, not a grant: asset-retention and reporting conditions apply, and NOHFC may require reimbursement if funded assets are sold after project completion. Combined government funding from all sources cannot exceed 50% of eligible project costs. Retail, accommodation, food service, and wholesale businesses are not eligible. Funding is limited and not all qualifying projects receive support. Confirm your specific eligibility and the current intake status with NOHFC before applying.

What Anthropic actually announced

Claude Sonnet 5 launched June 30, 2026, as what Anthropic itself describes as a capability upgrade over the previous Sonnet 4.6 model, with the largest gains in coding and agentic tasks — the kind of AI that does not just answer a question but carries out a multi-step job: reading a document, deciding what to do with it, calling other tools, and producing a finished output. Getting that level of reliability out of Claude has generally meant paying Opus-tier prices: $5 per million input tokens and $25 per million output tokens for Opus 4.8. Sonnet 5 launched at $2 and $10 — a rate available through August 31, 2026 — putting Opus-class agent behaviour within a Sonnet-sized budget for the first time.

The number in the documentation, not the press release

Here is the part that did not make most of the coverage. Anthropic's technical documentation for Sonnet 5 states plainly that the model uses a new tokenizer, and that the same input text produces approximately 30% more tokens than on Claude Sonnet 4.6. The company is direct about the effect: per-token pricing is unchanged, but because the same text produces more tokens, the cost of an equivalent request can differ. In plain terms, the rate per token is lower, but a real document now gets chopped into more billable pieces. What that does to your actual invoice depends on what you are feeding the model — Anthropic does not promise it nets out in your favour.

There is a second detail worth noting. The $2/$10 pricing is introductory. On September 1, 2026, it reverts to $3 per million input tokens and $15 per million output tokens — which is exactly what Sonnet 4.6 already costs today. Anthropic frames Sonnet 5 as a capability upgrade over Sonnet 4.6 at the same standard price, not a lasting discount. The window where it is genuinely cheaper than the outgoing model is about two months, not the new normal.

Why this is worth an afternoon for a Northwestern Ontario operation

None of this is really a story about Anthropic. It is a story about the price of AI that does actual work — not just chat — dropping into a range a small operation can justify testing. A forestry consultancy triaging inspection reports, a Thunder Bay accounting office drafting client correspondence, a municipal office processing records requests, a healthcare administrator managing intake paperwork: these are exactly the multi-step, document-heavy jobs agentic AI is built for, and until this week the credible version of that tooling ran on Opus-tier pricing that made a pilot hard to justify against a real budget. That barrier just moved.

• Get a real cost estimate, not a vendor's benchmark number — run your own documents through a pilot and measure the token count yourself. The 30% tokenizer increase means a generic per-token comparison will mislead you.
• Treat August 31, 2026 as a decision deadline, not just a launch date. A pilot worth doing at introductory pricing needs to run and get evaluated before the rate reverts.
• Do not commit a production workflow to Opus-tier spending you have not first tested at Sonnet-tier pricing — the capability gap between the two just narrowed.
• If a pilot pencils out, that project cost is the kind of eligible expense some funding programs are built for — see below.

A practical first step

Before scoping anything bigger, run one real task through Claude Sonnet 5 with your own material — an actual batch of client documents, inspection reports, or intake forms, not a demo. Measure the token count and the resulting invoice yourself, weigh it against what the same job costs you in staff time today, and do that math before September 1. If the economics work, NOIC's BBAA grant (up to $20,000, at up to 50% of eligible costs) and FedNor's RAII program both exist to help offset exactly this kind of AI adoption cost for Northwestern Ontario businesses — confirm current eligibility and intake status directly with each program. A local implementation partner can also help scope a pilot and get the token math right before you commit a workflow to it.

The discount is temporary and the token count is not what it used to be. Before building any workflow around Claude Sonnet 5, measure the actual cost against your own documents — not the per-token sticker price — and do it before introductory pricing ends August 31, 2026.

What BSP funds

BSP is built for projects that help an existing business scale and become more productive: adopting new technology, automating or digitizing operations, expanding capacity, and entering new markets. FedNor specifically notes it is well-suited to technology adoption and digital transformation projects — which is where an AI or systems investment tends to fit. The point of the program is a measurable step up in the productivity or growth of a real, operating business, not a first-time startup idea.

Who qualifies

Applicants must be incorporated small and medium-sized enterprises, and Indigenous (First Nation, Metis, Inuit) businesses and organizations are eligible. The project's benefits must accrue to Northern Ontario. You are normally expected to demonstrate a sustainable business model supported by at least two years of operations, plus the financial and managerial capacity to carry the project. Retail and service-based businesses are not considered under this program — it is aimed at businesses producing or scaling something beyond local retail or service delivery.

How much you can get — and the repayable catch

The contribution is up to 50 percent of eligible and supported project costs, to a maximum of $500,000 per project. Total government funding is capped at 75 percent of supported eligible non-capital costs and 50 percent of eligible capital costs. Two things matter here. First, the awarded amount is based on the minimum required for the project to proceed — FedNor funds the gap, not the wish list. Second, it is repayable: repayment terms are set when the contribution agreement is developed and normally begin no later than one year after the project is complete. Treat BSP as low-cost, patient financing for a growth project, not free money.

Is BSP a fit for an AI or systems project?

Often, yes. A project that automates a manual process, digitizes operations, or adds capacity through technology is exactly the kind of productivity improvement BSP exists to support — and up here, FedNor funding is a regional advantage a southern competitor cannot access. The test is whether the work makes an established Northern Ontario business measurably more productive or able to scale. If it does, and you can carry your share of the cost, it is worth a conversation before you commit your own capital.

How to apply

FedNor asks you to call 1-877-333-6673 to discuss your proposal with a FedNor Officer before applying, or to start through its online application portal. Because the budget is limited, that early conversation matters — an officer can tell you whether your project and timing are a realistic fit before you invest in a full application. As with every program in our funding radar: confirm eligibility, amounts, and current intake with the program itself, because these change.

Is BSP a grant?

No. BSP is a repayable contribution, not a grant. You receive funding up front toward eligible costs, then repay it on terms set in the contribution agreement. That still makes it attractive — it is patient, low-cost financing tied to a growth project — but plan for repayment, not a giveaway. If you need non-repayable funding, other programs in the radar (or a tax credit like SR&ED) may fit better.

Can a brand-new business apply?

Generally no. BSP normally expects a sustainable business model supported by about two years of operations, so it is aimed at established businesses rather than pre-revenue startups. Early-stage founders in Northwestern Ontario should look at startup-focused programs and accelerators instead — several are listed in the funding radar.

What RTRI funds

RTRI is designed to help established Northern Ontario businesses adapt to disruption caused by trade tariffs — whether that means reducing exposure to tariff-affected inputs, entering new markets, or strengthening domestic supply chains. Eligible activities confirmed by FedNor include digitization, automation, and technology integration; market diagnostics and expansion; strategic alliances and supply-chain optimization; domestic supply-chain strengthening and reshoring of production or R&D; and business support and market development services. Eligible costs cover labour, materials, capital equipment, consultancy fees, and advisory expenses. Retroactive costs going back up to 12 months before your application date may be considered, provided they did not occur before March 21, 2025.

Who qualifies

Applicants must be incorporated SMEs located and operating in Northern Ontario — Indigenous businesses are explicitly included. FedNor sets three baseline requirements: a minimum of five full-time employees; viable operations for at least three consecutive years; and evidence of tariff impact, defined as either 25 percent or more of sales in tariff-affected markets, or demonstrable negative impact from tariffs. Priority sectors include manufacturing, steel, automotive, critical minerals, mining, forestry, clean technology, bioeconomy, and agriculture. Retail businesses and tourism operators are not eligible.

The funding structure: non-repayable up to $1M, repayable above it

RTRI has two tiers. For projects up to $1 million in non-repayable funding, FedNor covers up to 50 percent of eligible costs — the business contributes the other half. For projects needing more than $1 million, the amount above that cap can come as a repayable contribution at up to 75 percent of eligible costs, with repayment terms set in the contribution agreement. As with all FedNor programs, the contribution is calibrated to the minimum amount required for the project to proceed. Confirm current budget availability with FedNor before investing in a full application.

RTRI excludes retail and tourism businesses. Applicants must have at least five full-time employees and at least three years of viable consecutive operations, and must demonstrate a direct tariff impact. Contact FedNor before applying — the two-phase portal process and available budget affect timing. All figures and intake status should be confirmed with the program directly, as terms can change.

Does a technology or automation project qualify?

Often, yes — if the project is a direct response to tariff exposure. Digitization, automation, and technology integration is listed as an eligible activity by FedNor, which can encompass deploying AI tools, automating a production or logistics process, or rebuilding an operation that previously relied on tariff-affected inputs or markets. The test is whether the technology investment is a legitimate adaptation to tariff impact, not whether the technology itself is eligible in isolation. If your operation has measurable tariff exposure and technology is part of your response plan, a conversation with a FedNor Officer is the right next step.

Is RTRI a grant or a loan?

The non-repayable portion — up to $1 million — does not need to be paid back, making it function like a grant for projects within that ceiling. Any repayable funding above the $1 million cap works like a low-cost loan, with repayment terms set in the contribution agreement. That makes RTRI meaningfully more attractive for smaller projects than programs like BSP, which is fully repayable.

How long is RTRI open, and should you apply early?

The official end date is March 31, 2028, giving businesses a planning window through the current tariff environment. However, available budget and intake conditions can change before the official end date — if demand is high, effective intake can close earlier. Contact FedNor early to understand current intake status and confirm the program is still accepting applications in your sector before building a project timeline around the deadline.

How to apply

RTRI uses a two-phase application process through FedNor's self-service portal. Before applying, contact a FedNor Officer at 1-877-333-6673 to discuss your project — this step helps you understand whether your project and timing are realistic before you invest in a full application. As with every program in our funding radar: confirm eligibility, amounts, and current intake with the program itself, because these can change.

What the strategy actually added to FedNor RAII

The AI for All framework allocated $500 million nationally to expand the Regional Artificial Intelligence Initiative, delivered through Canada's six regional development agencies. FedNor is the agency for Northern Ontario, which means its RAII program — already running on continuous intake since Budget 2024 — received additional mandate and budget. The program's two pillars remain: Pillar 1 supports AI companies scaling from prototype to market (commercialization), and Pillar 2 funds established SMEs integrating AI into their operations (adoption). Eligibility continues to favour incorporated SMEs, not-for-profits, Indigenous communities, and municipalities. Retail and service-based businesses remain outside the program's intent. The $500M is split across all six RDAs — FedNor's Northern Ontario share is not separately published.

The new piece: BDC LIFT

BDC launched LIFT (Lead with Innovation and Focus on Technology) on April 24, 2026 — a $500-million loan fund for Canadian SMEs investing in AI, digital tools, and automation. It was incorporated into the AI for All framework when the national strategy launched in June. The key details, from BDC's own program page:

• Digital Transformation & AI path: covers Canadian AI tools, data infrastructure, CRM/ERP systems, and cybersecurity. Minimum $1 million in annual revenue. A BDC Advisory Services Plan is required before financing is approved.
• Productivity & Advanced Equipment path: covers automation, robotics, and machinery. Minimum $5 million in annual revenue. Eligible sectors include manufacturing, mining, construction, transport, agriculture, forestry, and engineering — sectors central to the NWO economy.
• Preferential rates apply when technology or equipment comes from a Canadian supplier.
• Loan amounts start at $25,000; the total depends on your revenue, project scope, and business profile.
• Principal payment deferral of up to two years is available.
• BDC has a Thunder Bay business centre at 973 Balmoral Street, Unit 201. Applications run through BDC (1-877-232-2269 or bdc.ca/en/solutions/lift).

The adoption gap this strategy is designed to close

The government's own data sets the baseline: about 12 percent of Canadian businesses currently use AI. The strategy's stated target is 60 percent by 2034 — and Canada sits well behind peer countries, which is partly why the money is moving now. For a business in the North, the competitive frame matters more than the target: a peer in mining services, forestry, or logistics that is integrating AI into its operations is likely reducing costs or improving throughput. The funding exists to close that gap faster, with a lower cash outlay, than waiting for the tools to become unavoidable.

How to use this if you run a NWO business

• If you have a concrete AI adoption project and at least $1 million in annual revenue: contact BDC Thunder Bay about LIFT first. The required advisory step is also the scoping work a good application needs anyway.
• If you have a project driving genuine productivity or scale outcomes and you are an incorporated SME: contact a FedNor Officer (1-877-333-6673) about RAII — the AI for All expansion gives the program a stronger mandate.
• If your annual revenue is below $1 million: LIFT's Digital AI path is out of reach. NOIC BBAA (up to $20,000, up to 50% of costs) remains the most accessible AI adoption funding for smaller NWO operators.
• Apply before you start the work. Both RAII and LIFT require the project to be prospective, not retroactive.

LIFT is a loan, not a grant: you repay it on agreed terms. RAII is a repayable contribution for for-profit businesses. Neither is free money. Eligibility is set by the program — confirm your specific situation with FedNor or BDC before building a project plan around either one.

The sovereignty angle beneath the strategy

The AI for All strategy is partly a response to what the June Fable 5 shutdown made concrete: Canada's businesses rely heavily on US-controlled AI infrastructure that can be switched off without notice. The strategy's emphasis on Canadian AI tools — LIFT's preferential rates for Canadian suppliers, a $700-million sovereign compute expansion — reflects a deliberate attempt to build a domestic layer with less foreign dependency. For most NWO businesses the practical programs matter more than the policy rationale, but it explains why the money is moving now and why the regional distribution is intentional.

OpenAI's first custom chip signals a cost shift coming for AI users

On June 24, OpenAI and Broadcom unveiled the Jalapeño, OpenAI's first custom-built AI inference processor, completed in nine months from initial design to manufacturing tape-out. The chip is designed to run large language models in production — the inference side of AI, which is what a business actually pays for each time it calls an AI service. OpenAI said the Jalapeño delivers significantly better performance per watt than current alternatives, with initial deployment planned for late 2026. As model providers move onto custom silicon rather than rented GPU clusters, the per-query cost of running AI tools is expected to decrease over time. For any NWO business currently weighing whether AI tools are worth the ongoing operational cost, the direction of travel on inference pricing is now clearer.

Confederation College approved to train General Machinists — trades apprentices no longer need to leave the Northwest

Ontario's Ministry of Labour, Immigration, Training and Skills Development approved Confederation College as a Training Delivery Agent for the General Machinist apprenticeship program (trade designation 429A) on June 26. The designation means tradespeople entering mining, equipment repair, manufacturing, and resource-sector roles can complete required classroom training in Thunder Bay rather than travelling south. College President Michelle Salo and Associate Dean of Trades and Apprenticeship Andrew Phillips confirmed the program is ready to deliver. For NWO employers in mining, forestry, and manufacturing that have struggled to retain apprentices through multi-year programs, local delivery removes a significant structural barrier to completion and retention.

Thunder Bay's national workforce campaign is live — the CEDC is recruiting tradespeople from across Canada

The Thunder Bay Community Economic Development Commission (CEDC) launched the 2026 'Work Here' campaign in June, a national marketing effort running through September aimed at attracting tradespeople and professionals to the region. The campaign — tagline 'Come for work. Stay for the lifestyle' — includes a dedicated site at workinthunderbay.ca and advertising on the GO Transit network in southern Ontario. For regional employers competing for talent against larger markets, the CEDC campaign does the initial brand-awareness work at a regional scale; individual businesses do not need to build the 'why Thunder Bay' argument from scratch.

Thirteen active wildfires in the Northwest as of June 25 — an elevated-risk season for outdoor operations

As of June 25, thirteen wildfires were active across the Northwest Region, with four uncontrolled and five new fires confirmed that day in the Red Lake and Nipigon districts. Fire danger ratings were running high to extreme across much of Northwestern Ontario. Wildfires affect NWO businesses directly — through work stoppages, road closures, and air quality impacts for outdoor and construction operations. The Ministry of Natural Resources publishes daily fire status and maps on the provincial fire portal; operations in at-risk districts should be monitoring current conditions.

Geraldton gets an $81.3M rebuild — the Ring of Fire gateway is being prepared for mining traffic

A formal groundbreaking on June 2 launched the Geraldton Main Street Rehabilitation Project, described as the largest single infrastructure investment in the Municipality of Greenstone's history. The $81.3 million project reconstructs nearly five kilometres of Geraldton's main corridor from Highway 11 to Highway 584, with a three-year completion timeline; the contract was awarded to Pioneer/Minodahmun Development LP, a First Nations-owned firm. Geraldton is a highway gateway community on the route designated to carry Ring of Fire mining traffic, and the rebuild is explicitly designed to prepare the corridor for that load. Combined with the Webequie Supply Road groundbreaking on June 25, Ring of Fire corridor infrastructure is now under active construction on multiple fronts simultaneously. For NWO contractors and service businesses, the aggregate scope and multi-year timeline of these projects represent sustained regional procurement worth tracking.

Traditional industrial robots are precise but brittle. They repeat the same motion, in the same space, under the same conditions — and stop the moment something shifts. What AI adds is perception and decision-making: a machine can read a camera feed, identify an object it was not explicitly programmed to recognize, and adjust in real time.

That matters for tasks with variability: sorting irregular materials, navigating a route that changes day to day, inspecting equipment in conditions that are not perfectly controlled. Those are exactly the tasks common in mining, forestry, and heavy logistics.

Where it lands first in the North

Northwestern Ontario’s industrial base — mining along the Ring of Fire corridor, forestry, rail and trucking logistics — sits in the sectors where AI-driven automation is most mature. Autonomous haulage in mining, drone-based inspection, and AI-assisted sorting and grading are in use in the broader sector, and adoption is moving toward Northern operations as it reaches Canadian producers.

• Autonomous haulage and load management in mining operations.
• Drone inspection for infrastructure and remote forestry blocks.
• AI-assisted sorting and grading in processing and logistics facilities.
• Predictive maintenance that flags equipment failure before it happens.
• Route optimization and load planning in long-haul trucking and rail.

The workforce question

Northern Ontario already faces labour shortages in trades, mining, and skilled operations. Automation does displace some repetitive or dangerous tasks. It also creates demand for people who can configure, maintain, and interpret these systems — roles that tend to pay better and carry less physical risk than the tasks they replace.

The honest picture is mixed. Not every displaced role converts cleanly into a new tech role for the same person, and reskilling is real work whose timelines need to match the pace of adoption. What is clear is that facilities adopting these tools will want workers who understand them, and operations that fall behind will have a harder time competing for contracts and capital.

A note on humanoid robots: the demos are impressive — bipedal robots that walk, handle objects, and do basic tasks in a lab. That is genuine progress. But commercial deployment in a real industrial environment — on a floor with other workers, in a Northern climate, on an unpredictable site — is still early and largely unproven. The gap between a controlled demo and reliable daily operation is wide. Watch the space, but do not plan around humanoid robots solving a labour problem in the near term. The automation that matters for NWO right now is purpose-built for specific tasks, not general-purpose.

What this means for businesses in the region

If you operate in mining services, forestry, or industrial logistics, the question is not whether automation will affect your sector — it is which tools are deployment-ready now, what they cost to integrate, and whether your workforce and systems can absorb them. For smaller operators, the entry point is often software-side AI — fleet telematics with predictive alerts, AI-assisted dispatch and scheduling, automated inspection reporting — before any physical robotics investment. Those are lower-risk first steps that build familiarity with the category.

Projects that introduce automation, improve productivity, or reskill workers may be eligible for funding through programs like the FedNor Regional Artificial Intelligence Initiative (RAII) or the Northern Ontario Heritage Fund Corporation (NOHFC). Eligibility depends on your project scope and timing — confirm directly with each program before building a plan around it.

The regional opportunity

The Ring of Fire development corridor — chromite, nickel, cobalt — is a long-cycle investment that will need modern extraction and processing methods to be viable. Operations competing for that capital will need productivity and safety profiles consistent with what major investors expect, and AI-assisted operations are increasingly part of that picture. Northwestern Ontario does not tend to adopt early for its own sake, and that caution is reasonable — but in sectors where the technology is already proven and competitors are deploying it, waiting too long is its own risk.

A note on statistics: projections about AI’s impact on employment vary wildly depending on the methodology, time horizon, and assumptions. Some point to significant displacement; others project net job creation. Neither camp has a clean track record predicting technology’s labour effects. Treat specific percentages or timelines in headlines with real skepticism — including the reassuring ones. This post does not cite specific forecasts because none are reliable enough to anchor an honest piece. General information, not career or economic advice.

What the more careful research does suggest: AI tends to automate tasks, not whole jobs. A job is a bundle of tasks. Some — drafting routine emails, scheduling, sorting data, answering common questions — are being absorbed now. The rest of the role remains. Over time, what a job involves shifts: the title may persist while the daily work changes considerably.

The roles under the most pressure are those where repetitive, rules-based tasks make up the bulk of the work. The roles that tend to hold or grow are those where judgment, relationships, and local knowledge matter — things AI handles poorly.

Why the Northern Ontario picture is different

NWO does not face the same labour dynamics as southern urban centres. Many employers here — in trades, healthcare, retail, food service, transportation — already cannot fill the roles they have. The problem is not too many workers; it is too few.

In that context, AI’s most realistic near-term contribution is capacity extension. A three-person admin team with an AI assistant can handle the volume a five-person team used to. An owner-operator can answer after-hours inquiries automatically without hiring a coordinator. A small clinic can cut the time staff spend on scheduling and documentation. None of that is a layoff — it is doing more with the same people, which is a meaningful shift for a region where labour is the constraint.

Reskilling is the practical lever

The people best positioned through this transition are those who learn to use AI tools in their field — not those who wait, and not those who assume the tools will replace them before they have tried them. That holds whether you are a bookkeeper, a service advisor, a property manager, or a marketing coordinator.

• Identify the most repetitive, time-consuming parts of your role or your team’s work.
• Find one AI tool that applies to that specific task — not a strategy, a specific tool for a specific job.
• Use it for 30 days and measure whether it actually saves time.
• Build from there. One habit compounds into fluency over a year.

For employers, the version of this is giving your team time and permission to experiment. Reskilling does not require a formal program; it usually starts with someone already curious, given room to figure it out and share what works.

Funding that may apply

Several programs support AI adoption and workforce development in the North. Confirm eligibility and current availability directly with each program before building a plan around it — criteria change.

• FedNor RAII (Regional Artificial Intelligence Initiative) — funds AI adoption projects for businesses and organizations across Northern Ontario.
• NOIC BBAA (Northwestern Ontario Innovation Centre — Building Blueprints for AI Adoption) — supports local SMEs implementing AI.
• NOHFC Workforce Development Program — may subsidize training costs and internship wages for businesses investing in staff development.

These programs exist because governments recognize Northern businesses face a real disadvantage if they fall behind on technology adoption. That is a signal worth taking seriously — and a practical reason to act now rather than wait.

AI will not leave the workforce unchanged. But the shape of the change in Northern Ontario — where labour is scarce and small teams are the norm — looks more like a capacity tool than a displacement wave. The workers and employers who engage with it thoughtfully, starting now, will be in a better position than those who do not. That is not hype; it is the reasonable read of what is actually happening.

Andrej Karpathy — a founding member of OpenAI and former Director of AI at Tesla — used the phrase "vibe coding" in early 2025 to describe a pattern he had noticed: writing software by feel, in natural language, accepting AI output without carefully reading it. He was not attacking it. He was naming something real.

The idea spread because it matched what many people were already doing. You describe what you want. The AI writes the code. You run it, see if it works, describe the next piece. You might not read a single line of the output.

What it genuinely unlocks

The barrier to building something small has dropped to near zero. That is a real change, and it is useful for a small business in Northwestern Ontario in specific ways.

• Prototyping an idea before paying a developer to build it properly — a working demo in hours, not weeks.
• Internal tools only your team uses: a simple inventory tracker, a form that emails you, a spreadsheet that calculates automatically.
• Automating a small repetitive task: pulling data from one place to another, formatting a report, sending a scheduled reminder.
• Testing whether a concept is worth building — spin up something rough, see if it solves the problem, then decide whether to invest.

None of these require you to understand the code deeply. They require you to understand your own problem clearly enough to describe it — a skill most owners already have.

Where it creates real problems

The risks are specific, not theoretical. AI-generated code can contain bugs it does not flag, handle data in ways you did not intend, or have security gaps a developer would catch on review. When it runs an internal spreadsheet, that is a minor inconvenience. When it handles a customer’s payment or stores their personal information, it is a liability.

The standing rule: do not ship code you do not understand — or that no one on your team can read — into any system that touches customers, handles money, or stores personal data. AI-generated code still needs a human who can read it to review it before production. This is not about distrusting AI; it is recognizing that the AI does not know what it does not know, and neither do you if you have not read what it wrote.

Maintainability is the other quiet problem. Code you did not write and do not understand is hard to fix when something breaks six months later — and something will break. For anything you plan to run long-term, you want someone who understands the codebase.

A practical frame for NWO businesses

Think of vibe coding as a drafting tool, not a finished product. Use it to move fast on low-stakes experiments, to scope what is possible before spending money, and to automate small things that do not need to be bulletproof.

When an experiment proves out and you want to build it properly — or when it needs to handle real customer data — bring in a developer to review, rewrite, or rebuild from your working prototype. The prototype is genuinely useful input; it is just not the production system. The tools most people use for this are Cursor (an AI-first code editor), Replit (browser-based, no install), and direct prompting through Claude or ChatGPT.

The honest version of the story: vibe coding lowers the cost of experimenting dramatically — it does not remove the need for careful engineering on systems that matter. Experiment freely. Ship carefully.

Think of a chatbot as a capable coworker at a desk who answers your questions. You walk over, ask, they respond, you walk away. ChatGPT, Claude, Copilot — when you are typing back and forth, that is chatbot mode.

An agent is closer to someone you hand a project to. You say "book a meeting with every new lead who filled out a form this week, and send them a confirmation." It goes and does it — checking the form tool, reading entries, cross-referencing the calendar, drafting and sending — without you clicking each step. The agent uses tools: it can browse, run code, read files, call software, fill out forms.

That is the shift in 2026. The conversation interface is still useful, but the industry has moved hard toward agents that actually do things. Open-source agents like OpenClaw can control software on your computer to complete tasks you describe; workspace platforms are shipping agent features that string tasks together across email, drive, and calendars.

Why the difference matters for a local business

A chatbot mistake is low-cost: it gives you a draft you do not use, or a summary with a wrong figure you catch. The consequence is a wasted minute.

An agent mistake can have real consequences. If it sends the wrong email to a customer list, it is sent. If it moves the wrong files, they are moved. The more an agent can do, the more a bad run can cost. That is not an argument against agents — it is an argument for setting them up with care.

Agents need guardrails before they need goals. Before you hand an agent access to your email, your CRM, or your files, decide what it can touch and what requires a human to confirm. Start with read-only access or single-task scope. A well-scoped agent that does one thing reliably is more valuable — and far less risky — than a broad one you are not watching.

Where each one fits

• Drafting a proposal, quote, or customer email: chatbot. Low risk, fast, easy to review before it leaves your hands.
• Summarizing a long document, a transcript, or a set of reviews: chatbot. You are asking for a reading, not an action.
• Answering your team’s questions about policy, specs, or past jobs: chatbot, ideally connected to your own documents.
• Following up with new leads, adding them to a CRM, scheduling a call: agent territory — but scope it tightly and review the first runs.
• Pulling weekly numbers from several tools into a report: agent territory, and one of the cleaner cases, because mistakes are visible before anyone acts on them.
• Responding to customers in real time without review: only worth considering after thorough testing on low-stakes interactions.

What most NWO businesses should do now

Master the chat tools first. Pick one — Claude, ChatGPT, Copilot — and build the habit of using it for drafting, summarizing, and answering internal questions. This takes days, and the payoff is immediate. It also builds the judgment you will need to decide what an agent should and should not touch later.

Agents are worth exploring when you have a repetitive multi-step task that costs real time every week. The test: is this something a person currently does by following the same steps, in the same order, with the same tools, most of the time? If yes, it is a candidate. If it needs judgment that varies by situation, keep a human in the loop — and that is fine. The businesses that get the most out of agents are not the fastest movers; they are the ones who scope carefully, watch the first runs, and expand access only when the last step works.

The one-line version: chat responds, agent acts. Start with chat, earn your way to agents, and keep a human checking until you are confident in what it is touching.

Google used to charge a separate add-on for Gemini inside Workspace. In 2026 they rolled it into the base paid plans — about $2 per user per month over the old no-AI price. If your bill nudged up slightly, that is why.

For a business in Thunder Bay, Kenora, Dryden, or Fort Frances, this matters for a specific reason: most local operators already use Gmail for their inbox and Google Docs or Sheets for quotes, schedules, and records. The AI now sits inside the tools you open every day. No new service, no moving your data, no convincing staff to learn another platform.

What Gemini can do inside each tool

• Gmail — "Help Me Write" drafts an email from a short prompt; thread summaries catch you up on a long back-and-forth; smart reply suggests quick responses.
• Docs — generate a first draft from a prompt (a quote cover letter, a service description, a staff policy), then adjust tone, and ask questions about a document in plain English.
• Sheets — describe a formula in plain English and it builds it; ask it to analyze a range, flag patterns, or suggest a chart. Useful for a quote log, a job list, or an inventory sheet.
• Meet — automatic transcription during calls, plus a meeting summary with action items afterward, so a weekly staff call or client update leaves a written record without anyone taking notes.

Practical uses for a local operator

A trades business can draft a follow-up quote email in thirty seconds instead of staring at a blank screen. A small retailer can clean up a messy inventory list by describing what a column should calculate. A service business running jobs across a wide geography can use Meet summaries to keep everyone aligned without a separate notes process.

Google also shipped Workspace Studio, a no-code automation builder where you describe a workflow in plain English and it builds the steps across Gmail, Docs, Sheets, and other apps. That is the more advanced end — useful once the basics are running, not required to get value from the core features. An optional premium add-on (roughly $20 per user per month) raises limits for video and image generation and higher-tier automation; most small businesses will not need it to start.

One privacy note: treat AI drafts the way you treat autocorrect — review before you send. Do not paste confidential client information (financial records, legal documents, health details) into a prompt unless you understand how that data is handled under your plan. Google publishes data-handling terms for Workspace; read them, or have someone read them, before using AI features on sensitive files. The tools are useful; the oversight is yours.

How to check if you already have it

Open Gmail and look for a "Help me write" prompt when composing. Open a Google Doc and look for a sparkle or star icon in the toolbar. If those are present, Gemini is active on your plan. If not, check your Workspace admin console or whoever manages the account — you may be on a legacy or free plan. For most NWO businesses already on a paid plan, the AI is there; the gap is not access, it is knowing what to do with it on a typical workday.

Most AI tools you have used are reactive. You type something; they respond. That is a chatbot. An agent is different. It takes a goal, breaks it into steps, and carries those steps out — writing files, calling services, triggering commands — without a human confirming each one. OpenClaw sits in that second category.

The practical difference: you could tell a chatbot "write me an email about my Tuesday appointment" and paste the result yourself. You could tell OpenClaw to pull your calendar, draft the email, and send it. Whether that second version sounds useful or alarming probably depends on how much you trust the system and how sensitive your data is. Both reactions are reasonable.

A quick history of the project

The project launched under the name Clawd in November 2025 as a small open-source experiment. It was renamed Moltbot, then rebranded OpenClaw, with the current name announced on January 29, 2026. Growth since then has been fast — the GitHub star count crossed 350,000, and OpenAI announced it would sponsor the project. The code is released under the MIT license, meaning anyone can use, modify, or build on it without paying royalties or asking permission.

What it can actually do

• Automate file management on your local machine — sorting, renaming, archiving.
• Send emails and manage calendar events through connected accounts.
• Run shell commands and scripts, with whatever permissions you grant it.
• Call external APIs and web services — booking platforms, CRMs, data sources.
• Chain multi-step workflows that would otherwise mean switching between several tools.
• Extend its own capabilities through AgentSkills — community-built modules for specific tasks.

For a small operator in Thunder Bay, Dryden, or Fort Frances who is already comfortable running their own hardware — a shop computer, a home-office machine, a local server — this is a meaningful list. Automation that used to need a developer or a paid subscription can be assembled from open-source pieces running on equipment you already own.

Why it has drawn government scrutiny

Governments and security researchers have flagged OpenClaw for review. The concern is not unique to OpenClaw — it applies to any agentic system that executes commands: when software can take actions on your machine, the question of what it can access, what it can send out, and who controls it becomes important. A misconfigured agent with access to sensitive files and an outbound connection is a real attack surface.

Security caveat — read this before installing. OpenClaw runs commands on your machine with whatever permissions you give it. If you grant it access to your files, email, and shell, it can read and act on all of that. For a non-technical user, that is a serious risk. Do not connect it to accounts holding sensitive client data, financial records, or credentials unless you understand exactly what you are granting. Run it in a sandboxed environment first. The open-source model means the code is auditable — but you have to be willing to audit it, or trust someone who has.

Where it could genuinely help

The honest use case for a Northwestern Ontario business is narrow but real. If you are tech-comfortable, already manage your own hardware, and have a specific repetitive workflow — pulling weekly reports, syncing data between two tools, sending templated follow-ups — OpenClaw can automate it without a monthly fee. You own the hardware, the data, and the workflow.

What it is not, right now, is a hands-off system for non-technical users. It needs configuration, an understanding of the permissions you grant, and some willingness to troubleshoot. The 100-plus AgentSkills are community-built, so quality varies. Treat it as a power tool, not an appliance.

The open-source angle matters

The MIT license matters for one specific reason: no vendor lock-in. If the company behind a tool disappears or starts charging, an MIT-licensed codebase stays available. That is not true of most AI tools you pay for monthly. For operators cautious about depending on a vendor they cannot see or control, the open-source model offers a different kind of stability — though OpenAI’s sponsorship means a major commercial player now has influence over the project’s direction, which is worth watching.

Capability: name the job, then test

A model that is excellent at long-form writing is not automatically the right choice for structured data extraction, live customer conversations, or code. Name the job first — drafting, summarizing, analysis, conversation — then run five real examples from your business through each candidate, side by side. The right fit usually becomes obvious in ten minutes, and it is often not the one with the loudest launch.

The Canadian question: where does your data go

This is where a Canadian operator should slow down. Consumer tiers and business tiers handle your data differently — what is stored, where, and whether it can be used to train future models. Before you put business or customer information into any of these tools, check the provider’s data-residency and training-use terms. For public bodies the bar is higher still: an Ontario municipality is bound by MFIPPA, which shapes what can be put into a third-party tool at all.

Access is not guaranteed

There is a risk most reviews skip: you do not fully control whether a US-based AI provider stays available to you. In June 2026, a US national-security directive forced Anthropic to cut access to its newest models for foreign nationals — Canadian users included — within days of launch. Everyday models stayed available, but the lesson is concrete: a decision in Washington can change Canadian access overnight, with no input from you. Do not build a critical workflow that only works on one provider; keep a fallback.

The practical rule: choose by task fit and data handling, not by this month’s leaderboard. Keep customer personal information out of consumer tools, and never let one provider become a single point of failure for something your business depends on.

So which one

For most small-business uses, either capable general assistant will do the job — the deciding factors are task fit, data handling, and price at your volume, not a benchmark. Pick the one that wins on your own five examples, set it up with the right data discipline, and revisit the choice periodically rather than chasing every new release.

Start where you are losing money: the missed inquiry

The most expensive AI gap for a local business is not a missing chatbot — it is the call or message that never gets answered after hours, or while you are on a job. An AI responder that texts back a missed call and replies to a web form in seconds recovers leads you are already paying to generate. For a high-ticket local business, a single recovered job can cover a year of the tool. Start here; it pays for itself fastest.

Then: turn the work you already do into less work

• Draft follow-up emails, quotes, and cover notes from a few bullet points, then edit — faster than a blank page, and you keep control of the voice.
• Summarize long threads, documents, and call notes into a few action lines so nothing slips.
• Turn a finished job or a good review into a short social post — the work is the content.

Where AI does not belong yet

Anything that makes a real decision about a person, and anything involving confidential customer data in a public tool, stays off the table for now. AI is a drafting and triage assistant — strongest when a human signs off on the output, weakest when you trust it blind.

One rule that keeps you out of trouble: never paste customer personal information or anything confidential into a public AI tool. If a workflow needs real customer data, that is a job for a properly set-up business system, not a free chatbot.

How to pay for it

You may not have to pay full price out of pocket. For an eligible Northwestern Ontario business, AI adoption is exactly what programs like NOIC BBAA (up to $20,000, up to 50% of eligible costs) and FedNor RAII exist to fund. The catch is to apply before you start the work, and to scope a concrete project — confirm eligibility with the program first.

Pick one move, run it for the quarter, and measure whether it saved time or made money. That beats adopting five things at once and keeping none.

The short version

• BBAA (Building Blueprints for AI Adoption, NOIC) — up to $20,000 at up to 50% of eligible costs, built for growth-oriented for-profit SMEs adopting AI. Starts with a 30/60/90-day plan.
• RAII (Regional Artificial Intelligence Initiative, FedNor) — up to 50% of capital costs and 75% of non-capital costs, aimed at incorporated SMEs and economic organizations that can show a productivity, scale, or skilled-jobs gain; retail and service-based businesses will not be considered. Continuous intake.

Side by side

• Who it is for — BBAA: everyday for-profit NWO SMEs putting AI to work. RAII: incorporated SMEs outside retail and service.
• Amount — BBAA: up to $20,000 (up to 50%). RAII: up to 50% of eligible costs; the project maximum varies, so confirm it.
• First step — BBAA: a 30/60/90-day AI Adoption Plan before the application. RAII: a concrete, scoped project you can show drives productivity or scale.
• Common exclusions — BBAA excludes distributors, resellers, and retail-only businesses; RAII will not consider retail and service-based businesses; both exclude projects already underway. Apply before you start the work.

Which one is you

If you run a typical Northwestern Ontario business — a shop, a clinic, a trades company, a local service — and you want to put AI to work in your operations, BBAA is usually the more natural fit, and the 30/60/90 plan doubles as the scoping you needed anyway. If you are an incorporated company outside the retail/service exclusion with a project that clearly drives productivity, new capacity, or skilled jobs, RAII is worth a direct conversation with FedNor. The honest dividing line is the exclusion list: RAII is not the universal answer it sometimes gets pitched as.

Eligibility, amounts, and intake timing are decided by each program, not by us. Treat this as a starting map and confirm your specific situation with NOIC and FedNor before you build a plan around either one.

Either way, the work that wins funding is the same: a concrete project, a rough cost, and a clear line on the gain it drives. That scoping is also what makes the build itself succeed — which is the whole logic of funding a build with a grant.

What actually happened

On June 12, 2026, the US government issued a national-security directive requiring Anthropic to shut off access to Fable 5 and Mythos 5 for all foreign nationals worldwide — three days after the models launched. On June 26, the US revised the Mythos 5 restriction to restore limited access for a list of approved US entities and Anthropic’s own foreign-national staff. Fable 5 remained fully suspended worldwide, and Canadian users stayed blocked from both. Anthropic stated it believed the government had cited a narrow, non-universal jailbreak and that it was working to restore access — with no timeline given.

Will it reopen?

Honestly, no one outside the process knows. The only restoration so far was narrow and scoped — limited Mythos 5 access for approved US entities — and Fable 5 has not come back at all. Anthropic disputes the basis and says it is working on it, but has not given a date. The safe reading is that reopening is uncertain, not imminent, and that even the partial restore was defined by nationality and approval — which is itself the real story.

The bigger shift: AI as governed infrastructure

Set the specifics aside and look at the shape of it. A government reached in and switched off a private company’s most capable product, scoped by the nationality of the user, on national-security grounds, within days of release. That is how states treat controlled, dual-use technology — advanced chips, encryption, aerospace parts — not how they have treated software. One event is not a regime, and it would be a mistake to overstate it. But it is a clear signal of the direction of travel: the frontier models are starting to be governed like strategic infrastructure, which means access can be conditional, national, and revocable.

Why this lands differently in Canada

Most Canadian businesses run on AI models built and controlled in the United States. That was a convenience until June; now it is a visible dependency. A decision in Washington can change what a Thunder Bay company can use overnight, with no Canadian input and no notice. That is a continuity and sovereignty question, not just a tech preference. Canada’s own response is already on the table: the June 2026 "AI for All" strategy commits roughly $2 billion over five years, with a dedicated regional initiative that explicitly reaches Northern Ontario — partly an effort to build domestic adoption and capacity so the country is not purely downstream of others’ models.

What a Northwestern Ontario business should actually do

• Do not build a critical workflow that only works on one provider’s model. Keep a fallback you have actually tested.
• Prefer setups where you can swap the model underneath without rebuilding the whole system — portability is now a feature, not a nicety.
• Keep your data handling provider-independent, so changing models is a configuration change, not a crisis.
• Treat model access as a live variable to monitor, the way you would watch a key supplier — not a settled fact.

The practical rule: treat model access like any other critical supplier — have a second source, and never let one foreign decision become a single point of failure for something your business depends on.

This is exactly the kind of shift this hub exists to track: a global AI story with a specific, practical edge for the Northwest. We will follow the access and policy moves as they land, with the Canadian and regional angle the national coverage skips.

Ring of Fire access road clears final environmental review

On June 25, Webequie First Nation announced formal approval to proceed with the Webequie Supply Road — a 107-kilometre two-lane all-season road connecting the community to the Ring of Fire mineral deposits in the James Bay lowlands, with 31 water crossings including six bridges. Ontario says the road will open by November 2030, four years ahead of the original schedule. Chief Lorraine Whitehead noted the project "has always been about more than access — it is about creating opportunity." For Thunder Bay, which is the established services and logistics hub for the Ring of Fire corridor, construction procurement and professional services are beginning to move.

Canada's 'AI for All' strategy commits $2B — with Northern Ontario explicitly in scope

Prime Minister Carney launched the national AI strategy on June 4, committing roughly $2 billion over five years. Targets include 250,000 new AI-related jobs and a rise in business AI adoption from 12% to 60% by 2034. A dedicated $500 million regional AI initiative is explicitly designed to push capacity beyond the Toronto-Montreal-Edmonton research triangle. Minister Patty Hajdu, responsible for FedNor, cited Northern Ontario workers directly. For NWO businesses, the strategy's regional funding flows through FedNor's Regional AI Initiative (RAII), which is already accepting rolling applications — this strategy gives that program a clearer mandate and makes continued funding more likely through the year.

US export controls suspended Anthropic's Fable 5 globally — Canadians still blocked

On June 12, the US government issued a national security directive requiring Anthropic to shut off access to its newest models — Fable 5 and Mythos 5 — for all foreign nationals worldwide. The order arrived three days after the models launched. On June 26, the US revised the Mythos 5 restriction to restore limited access for a list of approved US entities and Anthropic's own foreign-national staff, but Fable 5 remains fully suspended worldwide and Canadian users remain blocked from both. Anthropic stated it believes the government cited a narrow, non-universal jailbreak and that it is working to restore access — no timeline has been given. The incident is a concrete illustration of the risk of building critical workflows on US AI infrastructure: a government decision in Washington can cut Canadian access overnight, with no input from affected users.

Federal government initiates fast-track for $26B nuclear waste repository near Ignace

On June 24, the federal government announced it is initiating the process to designate the Nuclear Waste Management Organization's Deep Geological Repository as a project of national interest under the Building Canada Act. The DGR would be constructed 650 to 800 metres underground near Wabigoon Lake Ojibway Nation and the Township of Ignace in Northwestern Ontario. The project is estimated at $26 billion. Public consultations begin in coming weeks, with a listing decision expected in fall 2026. The DGR would be one of the largest single infrastructure investments proposed anywhere in Northern Ontario, bringing decades of construction activity and skilled employment to the Ignace area.

North Star Air restores Fort Frances air link, starting July 6

North Star Air announced three-weekly service between Fort Frances and Thunder Bay beginning July 6, restoring a route lost in 2024 when Bearskin Airlines exited the region. Fort Frances is contributing $500,000 over two years to offset possible operating losses. Introductory one-way fares are $300 for the first week, with standard fares projected at $450 to $799. For NWO businesses with clients or operations in the Rainy River District, the restored link means same-day connections to Thunder Bay without a four-hour drive.

What it covers

• Up to $20,000, covering up to 50% of eligible project costs — AI software, consulting, and integration services.
• Aimed at growth-oriented, for-profit SMEs in Northwestern Ontario with a valid Canadian business number.
• Funding is limited and runs in intakes — confirm the current amount and whether applications are open with NOIC before planning around it.

The part most people miss: it starts with a plan

BBAA does not hand you an application first. You complete a 30/60/90-day AI Adoption Plan before the funding application. That sounds like a hurdle; it is actually the point. Scoping the project well is the hard part of any AI build, and a clear plan is what gets funded — a vague "we want to use AI" is a slow no.

Who it is for — and who it is not

It is built for growth-oriented for-profit SMEs. Distributors, resellers, retail-only businesses, and — importantly — projects already underway are outside its scope. That last one matters: apply before you start the work, not after.

Eligibility, amounts, and intake timing are decided by NOIC, and the funding is limited. Treat everything here as a starting map and confirm your specific situation with the program before you build a plan around it.

How to be ready

• A concrete project — the specific workflow AI would change and the outcome you expect.
• A rough cost estimate, since the funding is a share of eligible costs.
• The productivity or competitiveness gain it drives — the lens the program evaluates against.
• Then book the pre-application consultation and build the 30/60/90-day plan.

Answer the inquiries you are currently missing

The most expensive AI gap for a local business is not a missing chatbot — it is the call or message that never gets answered after hours. A simple AI responder that texts back a missed call or replies to a web form in seconds recovers leads you are already paying to generate. Start there; it pays for itself fastest.

Turn the work you already do into content

• Draft follow-up emails and quote cover notes from a few bullet points, then edit — faster than a blank page, and you stay in control of the voice.
• Turn a finished job or a good review into a short social post; the work is the content.
• Summarize long threads, documents, or call notes into a few action lines so nothing gets lost.

Use it as a thinking partner, not an oracle

AI is strongest as a first-draft and second-opinion tool: pressure-testing a price, outlining a proposal, listing what you might be forgetting. It is weakest when you trust it blind. Treat every output as a draft a human signs off on.

One rule that keeps you out of trouble: never paste customer personal information or anything confidential into a public AI tool. If a workflow needs real customer data, that is a job for a properly set-up business system, not a free chatbot.

Pick one of these, run it for the quarter, and measure whether it saved time or made money. That beats adopting five things at once and keeping none.

It is also widely misunderstood. The number gets repeated; the eligibility rarely does. Here is the structured version.

What the program offers

• Up to 50% of eligible costs covered (the government funding cap) — the specific project maximum varies, so confirm it with FedNor.
• Continuous intake — there is no single annual deadline, which means the constraint is readiness, not the calendar.
• Focus on SMEs adopting AI to improve productivity, scale operations, or create skilled jobs.

Who it is for — and who it is not

RAII is aimed at incorporated SMEs and economic organizations in Northern Ontario that can show a genuine productivity or capability gain from adopting AI. The honest caveat most pitches skip: FedNor states that retail and service-based businesses will not be considered. That does not mean a Thunder Bay shop is shut out of all funding — NOIC programs are broader — it means RAII specifically is not the universal answer.

Eligibility is decided by FedNor, not by us. Treat everything here as a starting map and confirm your specific situation with the program before you build a plan around it.

What to have ready before you apply

• A concrete project, not a wish — the specific workflow or system AI would change, and the outcome you expect.
• A rough cost estimate for the build, since the funding is a percentage of eligible costs.
• A short line on the productivity, scale, or skilled-jobs impact — this is the lens the program evaluates against.

The work is mostly in scoping the project well. That is the half most owners get stuck on, and the half worth getting help with — because a vague application is a slow no.

Start with the task, not the model

A model that is excellent at long-form writing is not automatically the right choice for live phone calls, structured data extraction, or coding. Name the job first — drafting, conversation, summarizing, analysis, voice — then evaluate models against that one job. "Best overall" is a marketing phrase; "best for this" is a decision.

The four trade-offs that actually decide it

• Capability — can it do the task well enough that a human only edits, not redoes?
• Cost — per-use price matters enormously at volume and barely at all for occasional use.
• Speed — a live customer interaction needs a fast model; an overnight batch does not.
• Privacy and data handling — where does your data go, and is that acceptable for what you are putting in?

Run a real test on your own work

The single most reliable step: take five real examples from your business and run them through two or three candidate models side by side. The right choice usually becomes obvious in ten minutes, and it is often not the one with the loudest launch. Re-check periodically — the landscape moves fast, and the best fit a year ago may not be the best fit today.

Resist model-chasing. The cost of constantly switching tools is usually higher than the marginal gain from this month’s top model. Pick a good-enough fit, ship, and only switch when a real limitation bites.

Where it genuinely helps

• Document-heavy workflows — summarizing reports, drafting routine correspondence, making long policy documents searchable.
• Service requests — triaging and routing resident inquiries so staff time goes to the cases that need judgment.
• Accessibility — generating plain-language versions, alt text, and translations to widen who can actually use a service.

Where it does not belong (yet)

Anything that makes a consequential decision about a person — eligibility, enforcement, hiring — should keep a human firmly in the loop. AI is a drafting and triage assistant in these contexts, not a decision-maker, and presenting it as more than that invites both error and a loss of public trust.

The local guardrails that matter

Ontario municipal bodies are subject to MFIPPA, which governs how records and personal information are handled and gives the public access rights to records. That shapes what can be put into a third-party AI tool and how. Procurement rules, records retention, and a clear human-oversight policy all apply before a public body deploys anything.

General information, not legal advice. Any public-sector AI use should be cleared against your organization’s privacy obligations (including MFIPPA), procurement policy, and records rules first.

The right first project for most public bodies is small, internal, and document-focused — prove the value and the controls on something low-stakes before anything touches a resident-facing decision.

The shape of it

• A non-dilutive cash contribution plus several weeks of office space and mentorship — confirm the current amount with NOIC.
• A full-time commitment from the founder — this is not a passive cheque.
• It runs in intakes/cohorts rather than continuously — confirm whether applications are open before you plan around it.

Who fits

Costarter is built for early-stage technology or scalable-innovation startups — ventures with growth potential beyond a single location. Established franchises, multi-level marketing, and localized retail are outside its scope. If you are a high-growth founder, that focus is a feature: the room is full of people building the same kind of thing.

A full-time accelerator is a real commitment. The strongest applications come from founders who can name what they will validate during the program — not just that they want the funding.

The readiness checklist

• A one-line description of the problem and who has it.
• Evidence of demand — even informal: waitlist, pilot users, paying customers, letters of intent.
• What the program would prove — the specific milestone you would hit.
• Your availability — confirm you can commit full-time before you apply.

If you have those four, you are most of the way to a credible application. If you do not, the gap itself tells you what to work on first.

The play is to close that gap deliberately. A custom AI or automation build is, in the eyes of these programs, exactly the kind of adoption they exist to fund. So instead of treating the build as a cash expense, treat the grant as its funding mechanism.

The four steps

• Find the right program. RAII for AI-adoption projects at scale; NOIC programs for earlier or broader cases. The program has to match your business type — that is the most common point of failure.
• Confirm eligibility before you fall in love with the plan. Talk to the program, or have someone who knows the criteria check, so you are building on a real yes and not an assumption.
• Scope and apply. Most of the work is turning a vague intention into a concrete project with costs and an expected outcome — the same scoping that makes the build itself succeed.
• Build what the grant funds. The funding is for the project, so the project has to actually get delivered. This is where the model either pays off or falls apart.

The honest limit: not every business qualifies for every program, and RAII in particular will not consider retail and service-based businesses. The funnel still works — it just means matching the business to the right program, not forcing one program onto every business.

Why now

The timing is the whole point. The federal AI-adoption money is flowing through Northern Ontario now, the programs have a mandate to spend it, and the businesses that move while it is available are the ones that get a funded system out of it. The window is open; it will not stay open forever.

Done right, this runs end to end — find and confirm the program, scope the project, then actually build the website, CRM, or AI system the grant funds. The grant is the mechanism; the working system is the point. Connecting the two is exactly what this hub exists to help with.