At a company in the GTA, the CFO opens an email forwarded by the CEO, with a link to the Regional Artificial Intelligence Initiative (RAII) and one line on top asking whether the company can apply by November 24. Whoever answers will have to check the project against what FedDev Ontario asks applicants to show, which the program publishes in its application guide.

An AI project could be ready to apply to RAII when it can show FedDev Ontario six things: a specific problem, the data behind it, results that can be measured and verified, a team that can deliver, the company's share of the cost and a responsible AI plan with a timeline. The checklist below tests each item against what the program asks for and says how to tell you are ready and what to do if you are not yet. For amounts, eligibility and the application steps, our guide to RAII in southern Ontario covers the program from the start.

At a glance

Early deadline

November 24, 2026, if your project and eligible costs start before March 31, 2027

Final deadline

January 19, 2027, 4:59 p.m. EST

Your share

At least 50% of eligible costs; future expected revenue and in-kind contributions do not count

Responsible AI plan

At a minimum, human oversight, accountability and cybersecurity, with an implementation timeline

One application at a time

A business can have only one application with FedDev Ontario at a time, across its programs

Which RAII deadline applies to your project?

The RAII deadline that applies to your project depends on when it starts. FedDev Ontario's when to apply page asks businesses to apply by November 24, 2026, if their proposed project and eligible costs start before March 31, 2027, and it sets January 19, 2027, at 4:59 p.m. EST, as the final deadline for the intake.

For a project that starts later, the final date is the one to work toward. A useful first step is to agree internally on the date the first eligible cost will be incurred, because that single date decides which deadline the team is working against.

A project already under way could still fit, since the program's application guide allows projects and project costs to begin up to 12 months before the application date. Anything spent before a formal decision is "at the applicant's own risk", in the agency's words, so the company carries those costs on its own if the application is not approved.

What does a RAII readiness checklist cover?

A RAII readiness checklist tests six things against what FedDev Ontario publishes. The second column comes from FedDev's guide and program pages. The last two are our own criteria at Adivor, built from more than 1,100 hours of AI work with Canadian companies.

Is your AI problem specific enough, and is the data ready?

An AI problem is specific enough for a RAII application when it can be described as a process inside the company, with the steps that will change and the outcome expected at the end. Question 28 of FedDev Ontario's application guide asks applicants to describe "the detailed steps required to successfully achieve your expected outcomes", a request that is hard to meet with a goal such as using AI in customer service.

A description that names the process gives the later questions something concrete to build on. In a quoting process, for example, it says how long it takes to answer a request, how many quotes go out with errors and who reviews them.

The data question follows from that description. Once the process is named, the next things to establish are where its records sit, who controls access to them and whether they are complete enough for an AI system to use. A gap here could belong inside the project instead of blocking it, since the program's list of AI Adoption activities includes "strengthening data quality, governance and architecture" alongside the implementation work.

A practical way to test this row is to ask the people who do the work to walk through the process step by step, naming each system they open and where the delays begin. If that walk-through fills a page, the problem is specific enough, and the page already points to the metric the next test needs.

What results should a RAII application be able to measure?

A RAII application should measure results that FedDev Ontario can verify. The program's assessment criteria look at jobs created, revenue growth and improvements to productivity, competitiveness and business growth. The guide's outcomes section requires every value to be "measurable and verifiable" and "as realistic and achievable as possible", and those values are what the assessment reads.

Productivity is the outcome the guide leaves most open, because it does not define a metric or explain how to measure one. The company chooses its own, and a figure chosen after the project starts is hard to verify.

What makes a result verifiable is a baseline measured before the project and the same way after it. It can be the hours a team spends on a task each week, the time it takes to answer a request or the share of documents that need rework. If no one can produce the current figure this month, the data is probably not ready either.

Who will deliver the project, and who can sign for the company?

The company applying to RAII has to show who will deliver the project and how the business is run. FedDev Ontario asks for biographies of key management and technical staff, a company structure diagram and, in question 8 of the application guide, an outline of "how your firm is managed and governed now and in the future".

A project passes this test when one internal lead owns it and the technical people named in the biographies match the work the plan describes. The guide's 41 questions, from governance to financial projections, ask for the same inside knowledge.

The application itself also comes from the company. The apply as a business page asks applicants to confirm that the company can enter into binding agreements and that the person submitting is authorized to bind it. FedDev also rules out consultants and other third parties as the authorized and financial contacts.

Outside experts can still deliver project work. If you are weighing one, our article on how to choose an AI consultant in Toronto lists the questions to ask any firm before you hire, including how consultant fees are treated under RAII.

Whoever builds the system, the company has to show it can use what gets built. The guide asks applicants to outline the steps they have taken or will take to manage and protect IP, and to show that they hold or will obtain rights to any third-party IP the project relies on. If the project is funded, keeping an IP strategy in place becomes a condition of the agreement, so a company building on licensed AI models or software should know whose terms govern what it builds.

How much does your company need to fund, and in what form?

A company applying to RAII needs to fund at least 50% of the project's eligible costs, since FedDev Ontario covers up to the other half, with the amounts and repayment terms set out in our RAII guide. The form of that share matters as much as the amount, because the program states plainly that "future expected revenue and/or in-kind contributions are not eligible forms of matched funding".

A project passes this test when the company's share is identified and its sources could be documented quickly. Applicants also report the source and status of any other funding.

If the project is approved, the company has to confirm all sources of its matching funding within 30 days of being notified. Before the budget goes into the application, it is worth asking finance to name where each part of the company's half will come from.

If your company already claims SR&ED (Scientific Research and Experimental Development) tax credits for AI work, FedDev Ontario's how funding works page asks applicants to exclude SR&ED-eligible costs from the application when possible, so those costs are best separated in the same review with finance.

What should a responsible AI plan include for RAII?

A responsible AI plan for RAII must describe, at a minimum, human oversight and accountability frameworks and the measures that address cybersecurity risks, according to question 34 of the RAII application. The same question asks for a timeline for implementation and an explanation of how the risks were identified.

The program's assessment looks for "clear, concrete measures" showing how the project and the business support the responsible development and use of AI, and question 34 asks the company to attest to that commitment once the project is approved. The application also needs the RAII Supplemental Form, even if every answer on it is no, and FedDev encourages adding the company's cybersecurity strategy and certifications.

A good place to start is to write down who approves what before drafting the plan itself: who decides which tasks the AI handles, who reviews its outputs, who can stop it and who answers when it makes a mistake. Those names give the plan its accountability section and its timeline, and they also show which gaps cannot be closed by November 24.

What if your project is not ready by November 24?

If your RAII project is not ready by November 24, the final deadline is January 19, 2027. What the program's pages do not explain is how FedDev Ontario treats a later application for a project whose eligible costs start before March 31, 2027. FedDev Ontario also accepts only one application from a business at a time, across all its programs, so an incomplete version filed in November could stand in the way of the complete one.

Before deciding, confirm with FedDev Ontario through its contact page and ask: "Our proposed project and its eligible costs start before March 31, 2027. If we apply after November 24, 2026, how is our application treated?"

Some of the open rows could become part of the project itself. "Conducting AI readiness assessment" is the first activity the program lists for AI Adoption projects, so a structured review of process, data and governance could fit within an eligible project, subject to FedDev's assessment.

Work done to prepare the application is treated differently. According to the how funding works page, "any costs related to the preparation of an application cannot be reimbursed".

What does an AI readiness assessment involve before you apply?

Before a RAII application, an AI readiness assessment checks the process you want to change, the data it relies on and the people who would own the project. Its purpose is to give each of the six tests above an answer the company can back with records.

The assessment can be run by the company's own team, by an independent outside adviser or by both together. It needs the process owner, someone from finance, someone from IT or data and the signing officer, since each test draws on what one of them knows. An outside adviser brings a method and a second reading of the process, and the company's people supply the records, figures and decisions.

It usually ends with a map of the process as it runs today, an inventory of the data sources with their gaps written down and a baseline for each result, measured before any work begins. It also names who makes the decisions, who oversees the system and who signs for the company.

With an early cost estimate and a short list of priorities, the company can decide which project to put forward and what still needs work before anyone drafts the application.

Where does Adivor's Custom AI Roadmap fit before you apply?

Our Custom AI Roadmap, part of the AI Strategy line, is the one-to-two-week version of this assessment that we use to close open rows before a company commits to an application. It maps the workflows behind the process you want to change and ends with three to five AI opportunities in priority order, each with a timeline and the people and systems it would need.

The roadmap comes before any funded project, while the company is still deciding whether to apply and for what, so we treat it as preparation work and recommend budgeting it outside the application.

In Adivor's work with more than 30 Canadian companies, including more than 20 implementations under the former Canada Digital Adoption Program (CDAP), the roadmap usually fills two parts of this checklist. The workflow mapping covers the problem and data rows, and the resource requirements give finance a first figure for the company's share.

With those rows filled in, the company has a scope its own team can describe in the application. From there, we build the AI work the roadmap puts first through our AI Autonomous Agents line, and AI Enterprise Training leaves the team with the prompts, workflows and usage policies to run it.

Frequently asked questions about RAII readiness

Talk to Adivor

Which rows are still open?

In a free 30-minute call, we go through the open rows with you. Bring the table filled in as far as you can and the date you expect the project's eligible costs to start.

Book your free consultation →

This article is general information and does not constitute legal, financial or funding advice. RAII terms are as consulted on FedDev Ontario's website on November 2, 2026. Program details can change, so confirm them with FedDev Ontario before you apply.