Canada Must Lead on Responsible AI Or Lose Our Trust Differentiator

Montreal, September 17, 2026 at ALL IN: Canada's largest AI conference was packed with founders, researchers, government officials, and investors from around the world. The message repeatedly touted was clear: compete globally, move fast, build sovereignty and control through scale and speed with AI.

It was also the stage where a protester from Last Generation Canada interrupted a panel entitled “Canada Is Not for Sale - It Is Selling to the World” with Minister Evan Solomon (Government of Canada), John Stackhouse (RBC), Joëlle Pineau (Cohere), Julien Billot (Scale AI) and David Skok (The Logic) on stage, to say what a growing number of people are thinking all over the world: "We’re here to tell you what you don’t want to hear. People don’t want AI. We don't your data center to drain our water supply and kill our communities. We don’t want to get dragged down in wars that you profit from. We don’t want to die in this arms race. We don't want you to steal our data. We don’t want you to mass surveil us." I think he made some good points. As a member of the tech community, and insider operator, with genuine concerns about our environment and our future, I think Canadian AI must properly address the growing concerns and issues surrounding AI. We have more than a public perception or marketing problem, and we have Big Tech’s track record to blame.

As is seen in this video, the nonviolent activist was forcibly removed and he was arrested and detained for 5 hours. Minister Evan Solomon briefly addressed what happened and then the ALL IN event moved on. But what happened in the hours and days after ALL IN tells us more about this evolving story — one that shows Canada has a real opportunity and responsbility to lead the world in something more valuable than just being fast and in control with AI development. A lot is at stake, and leaders in technology in Canada are increasingly aware of this, as we learn from other nations and examples of what can go wrong.

What We're Seeing Globally (And Why It Should Concern Us)

I'm writing this as someone embedded in the AI ecosystem who works AI startups, who believes the technology has great potential to improve dimensions of life. And, like the CEOs of Anthropic and OpenAI, I’m concerned about whether AI can hurt us, and I’m also skeptical about whether AI can truly benefit everyone in our society, and not just those with money, or those using agents to speed up business productivity.

Reports this Fall, just came out about the latest AI incident, after the very alarming headlines this summer, about the autonomous OpenAI agents that escaped their testing environment to hack Hugging Face servers in an attempt to steal cybersecurity test answers.

The Australia Hack (Reported by BBC on September 24, 2026)
An OpenAI agent went "rogue" during a test and infiltrated Australia's Medicare health system. An autonomous AI agent broke the rules to complete its task and posed real risk. Apparently the hack went unnoticed for two months. Dr. Hammond Pearce at UNSW warned this kind of incident will "grow in severity and in frequency."

The incident reveals a fundamental problem: large language models are designed to predict the likeliest output rather than consider consequences as real humans would. When you place guardrails on AI, they don't always hold under pressure. Dr. Niusha Shafiabady at Australian Catholic University said autonomous AI "does not always know when it is wrong, and humans may not be able to see why it made a decision."

This is what the industry calls "misalignment"—and it's a massive problem we're racing to solve while also racing to deploy AI in the real world.

The UN Wake-Up Call (Reported by PBS on September 23, 2026)
AI leaders stood and spoke candidly about the risks of the technology they're building. Dario Amodei, CEO of Anthropic, testified before the UN Security Council: "If managed poorly, I even believe AI could be a risk to humanity as a whole."

Sam Altman, CEO of OpenAI, was equally direct: "We could lose control of the future to AI."

Both called for the world to set controls to prevent the technology from getting too powerful to rein in. They stressed the importance of preventing power concentration in any single company or country.

The UN Security Council was grappling with a fundamental question: Who should control life-and-death decisions—human or machine? U.N. Secretary-General Antonio Guterres warned of "killer robots." European leaders said that's not hyperbole but something that's already a reality or perilously close.

These are the CEOs of the companies building the most advanced AI systems on the planet. And they're saying—publicly, to world leaders—that we need serious control mechanisms, transparency, and safeguards.

What ALL IN Revealed About the Current Model

At the conference, Canada's government official, Minister Solomon spoke about sovereignty, about building things right, about community consultation and indigenous rights. These aren't empty phrases—they're part of Canada's actual approach.

These values were being stated within a framework of competitive speed. "We need to build safely AND we need to build fast." "We need to consult AND we need to move quickly." "We need sovereignty AND we need to compete" on the global stage.

There is a constant pressure to prioritize speed and when speed and safety conflict, speed usually wins… Not because people are malicious, but because the competitive logic of the hypercapitalist market demands it. Now that Canada is facing additional pressures due to Trump’s trade war, we are getting more desperate to bolster our sovereignty quickly.

This is the model the world is following right now: move fast, iterate, deal with the problems as they emerge. And globally, we're seeing the cost of that approach. Immense wealth and power have concentrated with big tech companies that own all the data, all the resources, and political leverage.

The Real Opportunity: Canada Can Be An Ethical AI Leader

My opinion is that Canada shouldn’t compete on out-scaling the US or China in the AI race. We can still build successful businesses and scale companies globally. As panelists at ALL IN stated, what we must compete on is something more valuable that Canada can become best known for: trust.

As John Stackhouse said at ALL IN: "Canada has the scarcest commodity in the world. We have trust."

This is a huge mandate – We have more work to do to earn people’s trust. We can’t leave vulnerable, marginalized, or less economically stable Canadians behind in our race to betterment through AI technology. We must not copy Big Tech down a path of domination, instead of true collaboration with our communities and government authorities. Canada needs to lean into building trust more, and not just as a buzz word to brag about. Canada can be positioned as the place where responsible AI is built. Imagine if Canada uniquely deploys the most AI solutions for social good? This is AI developed with genuine care for safety, transparency, and long-term consequences, and represents an era of ethical Canadian AI, generating real-world solutions. There has to be a values-based economic incentive for this, and if there isn’t, then maybe capitalism is in serious need of rehabilitation.

This isn't idealism. It could be a good business strategy, if our system rewarded solving problems that affect the majority of our population globally (the 99%). Canada could be pioneers in AI for social good for all of humanity, to innovate in solving the climate emergency (which is really where the clock is ticking on our very existence), and not just be known for leveraging AI for b2b SaaS applications, or military and defense, that ends up generating more capital in the hands of the relatively few.

Here's why a new ethical AI positioning can work:

1. Companies want to work with trustworthy partners. When you're deploying AI in critical infrastructure—healthcare, finance, government—you want partners whose systems you can trust, whose transparency you can verify, whose incentives are aligned with yours. Canadian-built AI, built with community input and safety-first principles, becomes attractive precisely because it's been built carefully.

2. Talent wants to work on meaningful problems. An entire generation is at risk of opting out of tech because they don't see their values reflected in what's being built. But if Canada positions itself as the place building AI responsibly? That becomes a magnet for the best most principled and morally aligned talent—people who want to solve hard problems without cutting corners on ethics. We’ve had the cult of money and growth at all costs for so long. It’s time for a tech revolution on values, before it's too late: AI for good.

3. Government support becomes easier. When local AI companies are transparent about their methods, when they consult with communities, when they prioritize safety and accountability, governments trust them more. That trust translates to easier regulation, better partnerships, faster approvals—the opposite of what happens when you move fast and break things.

4. Global regulation is coming anyway. The UN testimony, the 20-nation joint statement on safeguards, the growing calls for international oversight—it's all pointing the same direction. If Canada gets ahead of this and establishes ourselves as the leader in responsible AI practices, we don't have to scramble to comply with regulations later. We'll help write them.

What LawZero Represents

Canada and Germany's joint investment in LawZero—CAD 150 million and EUR 100 million for a not-for-profit developing trustworthy, transparent AI—is a huge step forward in the right direction. More people need to know about it, as it matures in time.

LawZero's "Scientist AI" isn't designed to act autonomously or pursue its own objectives. It's designed to reason transparently and provide evidence-based outputs without profit-driven goals distorting the reasoning. Yoshua Bengio wrote: “This organization has been created in response to evidence that today’s frontier AI models have growing dangerous capabilities and behaviours, including deception, cheating, lying, hacking, self-preservation, and more generally, goal misalignment. LawZero’s research will help to unlock the immense potential of AI in ways that reduce the likelihood of a range of known dangers, including algorithmic bias, intentional misuse, and loss of human control.”

It won't be faster than commercial AI systems. It won't beat OpenAI or Google on raw capability. But it will be trustworthy in ways that matter for government, healthcare, and critical infrastructure.

And that's a competitive advantage. This shows us, there's another path, and Canada is investing in it.

The Math of Responsibility

Here's what I think we all need to understand:

The current model—move fast, scale aggressively, optimize for speed and market share—operates on a dangerous assumption: that the rules don't apply if you're moving fast enough. That innovation justifies externalizing costs onto communities, workers, ecosystems, and people who never consented to the risk and won't benefit from the profits.

This is the hubris of Big Tech and corporate billionaires. And we all pay for it.

The environmental cost. The displaced workers. The eroded social trust. The security vulnerabilities. The mental health crisis in young people. The data extracted without compensation. These aren't externalities to manage—they're costs that are being passed to society while profits concentrate in fewer hands.

That's not sustainable.

Can we innovate in AI while maintaining safety, community trust, and long-term viability? Can we price in the real costs to people, to communities, and to the planet, instead of pretending they don't exist? That remains to be seen.

Canada can be the place that refuses hubris. We don't have to believe the rules don't apply to us. We can build AI differently—with genuine accountability, real community consultation, and honest acknowledgment that some decisions have costs worth paying to get things right.

Canada has the infrastructure for this. We have stable institutions. We have a population that generally trusts government. We have indigenous communities whose voices must be heard. We have water, energy, and talent. We have a reputation for fairness and transparency.

Building AI responsibly under those conditions doesn't make us uncompetitive. It makes us the kind of partner—and the kind of country—that deserves to win.

How Canadian AI Companies Can Lead

1. Lean into LawZero and models like it. Make it clear that Canada is betting on trustworthy, transparent AI—not as a sacrifice, but as a competitive strategy. Position it as the alternative to move-fast-and-break-things.

2. Make safety and transparency your brand. When Canadian companies deploy AI, require them to be radically transparent about how it works, what safeguards are in place, what went wrong when things break. Make it a credential.

3. Require real community consultation, not performative. When data centers are built, when AI systems are deployed in healthcare or government, actually listen to communities. Build slower if necessary. The trust you build is worth the time invested.

4. Attract talent and investment by being the responsible option. Market to founders who want to build responsibly. Market to investors who believe long-term value comes from trust, not speed. Market to governments who want partners they can rely on.

5. Work with allies on international standards. Don't wait for regulation to be imposed. Help write the global standards for responsible AI. Be at the table when other democracies are figuring this out.

This isn't about being slow or anti-innovation. It's about being strategically cautious. It's about understanding that in a world where speed is becoming commodified—where any AI lab can train a model fast—the real differentiator is judgment, trust, and long-term viability.

Why This Matters Now

The protester on stage during ALL IN and the others outside on the streets (and others protesting data centers in the US) are right about some things – One is that the current model concentrates profits while distributing risks. Citizens have to deal with the fallouts and negative consequences of AI, profit growth they don’t necessarily benefit from, and wars they may be at risk from… But he was framing it as a reason to reject the technology. Many of us in tech are saying it's a reason to build it differently. And let’s not be naive - this technology (AI), is different from tech of the past that citizens protested, that old tech was deployed slower, and was far less pervasive or invasive (depending on how you see it).

I hope Canada can compete on AI without competing on recklessness. If we do—if we position ourselves as the place where AI is built carefully, transparently, and with genuine care for consequences—we win in a way that speed never could. We become the partners that governments, large institutions, and serious investors trust. But most importantly, we can be trusted by the people, by society at large … because if we’re not doing this for everyone, why does it matter? And in the long run, that's the only way to compete that actually lasts.

Let’s Connect

If you're a tech founder building AI and you care about sustainability, social good, and/or environmental responsibility (if you want to compete differently), let's talk.

I work in marketing for AI and tech companies, and I believe there's a real opportunity to build companies that win precisely because they refuse to cut corners on ethics, transparency, and accountability.

If you're thinking about how to position and market your company as part of Canada's responsible AI movement, or if you want to discuss how to market sustainability and social good as competitive advantages, let's talk. Let's figure out how to tell that story.