top of page
Search

Solomon Needs to move Faster.

  • Writer: Ryan James Purdy
    Ryan James Purdy
  • Apr 5
  • 5 min read
Image Credit: Christopher Katsarov, The Canadian Press / Ana Luisa O.J
Image Credit: Christopher Katsarov, The Canadian Press / Ana Luisa O.J

Key Takeaways


  1. Every sector in Canada is adopting AI right now without a single enforceable federal rule to guide them. That is not a trust problem. It is a governance failure.

  2. The Minister's office has produced committees, consultations, and voluntary codes. None of these are law. None of them give any organization in the country a compliance floor to build on.

  3. Everyone knows Canada's eventual AI legislation will borrow heavily from the EU AI Act. The frameworks exist. Stop studying them and start passing them.

  4. If the federal government is not going to govern this space, say so, and let the rest of us get on with building what is actually needed at the institutional level.


Evan Solomon, Canada's first Minister of Artificial Intelligence and Digital Innovation, likes to say that "technology moves at the speed of innovation, but adoption moves at the speed of trust." It is a good line. It is also wrong. Adoption is not waiting for trust. Adoption is happening right now, across every sector, in every province, without guardrails, without legislation, and without any indication that the federal government intends to provide either one.

The question is not whether Canadians trust AI. It is who, exactly, they are supposed to be trusting in the absence of any enforceable rules. The platforms? The vendors? The voluntary code of conduct that the Minister keeps referencing as though it were law?

I have been paying close attention to Minister Solomon's file since his appointment. His ministry does not exist as a standalone department. It sits inside Innovation, Science, and Economic Development Canada, a department whose mandate is to help businesses grow and compete. His 28-member AI task force drew criticism from over 160 signatories, including the BC Civil Liberties Association and PEN Canada, for prioritizing industry voices. The Citizen Lab's Ron Deibert, one of Canada's foremost experts on digital threats, refused to participate entirely. The 30-day public consultation attracted over 11,000 responses and was then summarized by an AI tool. The Minister's guiding phrase for regulation is "light, tight, and right."

Light on what, exactly? We do not have a federal AI law. We do not have mandatory reporting requirements for AI platforms. We do not have enforceable transparency obligations. We do not have a regulatory framework that any business, school board, hospital, or municipality can point to and say: this is what compliance looks like. What we have are committees, consultations, and the promise that a national AI strategy is coming.

Meanwhile, every other sector is expected to figure it out on their own.

This is not just an education problem or a technology problem. It is a governance problem that touches every part of the Canadian economy. Businesses cannot confidently invest in AI integration when they have no way of knowing whether what they build today will be non-compliant tomorrow. School boards are deploying AI tools in classrooms without a federal or, in most cases, provincial framework to guide them. Health care providers, financial institutions, and municipalities are all making adoption decisions in a regulatory vacuum, hoping that when the rules finally arrive, they will not have to start over.

The Minister talks about productivity. He compares Canada's AI researchers to hockey stars. He tells a story about his father being afraid to use a credit card online. These are fine rhetorical moves. But rhetorical moves are not legislation, and Canada's AI sector cannot run on charm. Every serious observer knows that whatever Canada eventually passes will draw heavily from the EU's AI Act. The frameworks exist. The models are available. The question has never been what to do. It has been whether this government has the will to do it.

I work in AI governance for K-12 education. For the past several months, I have been conducting public assessments of school boards across Canada, evaluating their AI governance readiness against jurisdiction-specific legal baselines. I score boards on accountability structure, reporting pathways, decision authority, policy maintenance, and AI transparency. The findings are consistent: institutions are doing their best to build governance frameworks from the ground up, but they are building on sand. There is no federal floor beneath them. Most provinces have not operationalized the legislation they already have. Ontario's Enhancing Digital Security and Trust Act explicitly names AI governance and children's digital information but has produced zero implementing regulations. British Columbia has no equivalent framework at all.

These boards are not waiting for trust. They are waiting for direction. So is every other institution in the country.

The Tumbler Ridge shooting brought the platform accountability question into sharp focus. OpenAI confirmed it had flagged the shooter's account internally months before the attack but determined the activity did not meet its threshold for law enforcement referral. Minister Solomon said he was "deeply disturbed" and that "all options are on the table."

I want to know which table. Because I know what is not on it. Education is not on it. And neither, it appears, is legislation.

What the country needs is not another committee. It is not another consultation. It is not another voluntary code that platforms can adopt or ignore at their discretion. What the country needs is law. Enforceable rules that create a floor everyone can build on. Mandatory reporting obligations so that when a platform flags violent content, there is a legal requirement to act, not an internal threshold that nobody audits. A regulatory framework that gives businesses, schools, and public institutions the certainty they need to adopt AI without the constant risk that they are building something that will be non-compliant the moment the rules finally arrive.

The Minister keeps telling Canadians that this is a "hinge moment." He is right about that. But hinge moments require decisions, not consultations about decisions. Canada needed AI legislation yesterday. At this point, we would settle for any legislation at all. The rest of us have work to do, and we cannot do it responsibly in a vacuum.

Minister Solomon, if your office is ready to move beyond committees and into governance, my work and my methodology are public. The invitation is open. But so is the clock.

References

Federal AI minister raises concerns over OpenAI safety protocols after Tumbler Ridge mass shooting - CBC News, February 21, 2026 https://www.cbc.ca/news/canada/british-columbia/federal-ai-minister-raises-concerns-over-openai-tumbler-ridge-shooting-9.7101279

OpenAI safety reps called to Ottawa after Tumbler Ridge, B.C., mass shooting: minister - CBC News, February 23, 2026 https://www.cbc.ca/news/politics/open-ai-summoned-ottawa-tumbler-ridge-9.7103281

Evan Solomon Wants Canada to Trust AI. Can We Trust Evan Solomon? - The Walrus, January 15, 2026 https://thewalrus.ca/evan-solomon-ai/

People's Consultation on AI - Civil Society Initiative https://www.peoplesaiconsultation.ca/

Enhancing Digital Security and Trust Act (EDSTA) - Ontario e-Laws https://www.ontario.ca/laws/statute/24e24

EU AI Act (Regulation 2024/1689) - European Parliament https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32024R1689

 
 
 

Comments

bottom of page