Liam Gill leads the Capital Program at MaRS Discovery District, the largest innovation hub in Canada. In 2024, MaRS-supported companies raised 13 percent of all venture capital deployed in Canada.
The cracks in Ontario’s foundation are no longer just metaphorical, they are visible in our crumbling overpasses and felt in the quiet desperation of our hospital hallways. Ontario’s public services are at a breaking point; our roads and bridges need $34.7 billion to repair existing damage, the average wait time for emergency room patients to be admitted to a hospital is 18.3 hours, and only half of Grade 6 students meet the provincial math standard. These aren’t just line items on a provincial budget, they are deteriorating metrics of our collective quality of life.
AI could present a new approach to solving these challenges.
Ontario has the talent, the companies, and the need. What it does not yet have is a procurement system capable of bringing those pieces together.
Ontario has 140 emergency rooms; If we want to reduce wait times by 25 percent, we can either build 35 more emergency rooms to increase capacity by 25 percent or improve the efficiency of existing ones by 25 percent. This is the promise of AI: unburdening doctors, teachers, and other frontline workers by streamlining their administrative tasks so their specialized expertise can be directed toward healing patients or supporting students. We don’t need to wait a decade to train more staff or spend billions on new infrastructure to see meaningful improvement.
If Ontario wants to lead in AI, it has to become a first customer, not just a talent factory. We are home to what might be the strongest AI talent pools in the world, and global companies actively recruit our best minds to solve their own challenges. And yet, while Vienna tackles housing affordability with an AI-powered building permit review system or Sweden reduces unemployment as AI matches job seekers with open positions, Ontario’s public services rarely adopt the world-class AI built here.
I’ve personally worked with dozens of Ontario startups whose innovations could improve our public services. But too many founders have learned that if they want adoption in Ontario, they must first prove their technology elsewhere, usually in the United States.
Consider health care. One in five Ontarians doesn’t have a family doctor, and existing physicians spend an average of 19 hours a week on paperwork. A startup I supported through the MaRS Capital Program, founded here in Toronto, built specialized AI agents for health-care organizations, streamlining operations so clinicians can spend less time on faxes, calls, or emails. This can help existing providers to serve up to 15 percent more patients. After launching in Toronto, the company found that healthcare providers in Ontario were burdened by a complex procurement system. With American customers able to adopt the technology more quickly with fewer barriers, the company reincorporated in the United States.
The same pattern exists in infrastructure. Ontario’s transit and municipal assets continue to age faster than we can repair them. ConeLabs, based in Kitchener, uses AI to generate high-fidelity 3D models of bridges, buildings, and other critical assets from phone or drone images. Its technology allows engineers to detect structural problems earlier, before lane reductions, closures, and costly emergencies occur on critical assets like the Gardiner Expressway. But its strongest traction is coming from American customers and investors.
Education is facing similar pressures. Ontario has lost 5,000 teachers since 2019, increasing the burden on classroom educators. VibeGrade, another Toronto-founded company now backed by Microsoft and Y Combinator, uses AI to help teachers deliver faster, more actionable feedback on student writing to reduce time spent grading. Its product could ease workloads and support Ontario’s teachers. Instead, in 2025, both founders relocated to San Francisco where schools and districts have a clearer process for piloting and adopting new technologies.
RELATED: Canada will get a national summit in October to back procurement of homegrown tech
These are not isolated stories, they are three examples of a much larger pattern. Ontario is producing some of the world’s most promising AI solutions for health care, infrastructure, and education, precisely the public systems under the greatest strain. But without a clear pathway for governments and public institutions to test, validate, and adopt new technologies, these companies scale elsewhere.
The province’s current procurement process, built for large, predictable purchases, was never designed to evaluate emerging technologies or run pilots that demonstrate whether new solutions work in real-world settings. As a result, even when Ontarians build tools that can reduce emergency room backlogs, prevent infrastructure failures, or support teachers, there is no straightforward way for them to enter the public system.
Other jurisdictions have already recognized this gap. The United Kingdom, Singapore, and the United States have modernized procurement frameworks that allow governments to test and validate new technologies in weeks or months, not years.
If Ontario wants to lead in AI, it has to become a first customer, not just a talent factory.
Ontario should take two steps. First, provide more power to frontline workers like doctors and teachers by creating clear guidelines for them to initiate controlled pilots of new technologies with independent evaluation, clear privacy requirements, and the ability to scale what performs and stop what doesn’t. Second, create a single, province-wide pathway for startups to share their innovations with the provincial government to be centrally rolled out and piloted by public institutions.
Ontario has the talent, the companies, and the need. What it does not yet have is a procurement system capable of bringing those pieces together. If the province builds that pathway, it will help Canadian innovations scale, creating jobs and economic opportunity while strengthening the public services that we depend on every day.
The opinions and analysis expressed in the above article are those of its author, and do not necessarily reflect the position of BetaKit or its editorial staff. It has been edited for clarity, length, and style.
Feature image courtesy MaRS Discovery District.