On March 9th, AI Governance and Safety Canada Founder and Executive Director Wyatt Tessari L’Allié’s testified before the House of Commons Standing Committee on Industry and Technology (INDU) for their study on Opportunities, Risks, and Regulation of AI in Canada’s Strategic Industries.
It was the second appearance at INDU for Mr. L’Allié, the first being in January 2024 during the study of the AI & Data Act. This time the backdrop was the recent breakthroughs in AI agents producing a spate of loss of control incidents.
A separate organisation, Control AI, picked up on the testimony and posted a version on Instagram that reached over 2.3M views and 113,000 likes: https://www.instagram.com/ai_ctrl/reel/DWB-I6cgbqG/.
The opening statement is available at: https://www.youtube.com/watch?v=ZlHlpMpu8ug
And the full hearing is available at: https://parlvu.parl.gc.ca/Harmony/en/PowerBrowser/PowerBrowserV2/20260310/-1/44552
Full text:
Monsieur le Président, mesdames et messieurs les membres du comité, je vous remercie de l’honneur de m’avoir invité.
Gouvernance et Sécurité de l’IA Canada, est un organisme à but non lucratif non partisan et une communauté de personnes à travers le pays. Notre point de départ est la question “Que pouvons-nous faire au Canada, et à partir du Canada pour s’assurer que l’IA avancée soit sécuritaire et bénéfique pour tous?”. Depuis 2022, nous fournissons au gouvernement fédéral des recommandations de politiques d’intérêt public, telles que nos soumissions sur le projet de loi sur l’IA et les données, et nos multiples témoignages aux comités parlementaires.
Two years ago, in the context of the AI & Data Act, I testified before this committee that while early forms of AI like facial recognition and chatbots require some regulation, there were much more powerful forms of AI on the horizon that Canada needed to get ready for.
We made the case that certain AI capabilities pose an unacceptable risk because they could lead to dangerous weaponisation or loss of control scenarios. Systems that, without the instruction or authorisation of their users, can detect and evade monitoring, rewrite their own code, make copies of themselves, spawn other AI systems, commandeer resources, or refuse shutdown.
In the last few weeks, a major jump in AI capabilities has produced such systems. We have now entered a new paradigm in AI, the era of AI agents.
Unlike chatbots that simply respond to a prompt, AI agents are systems that can take actions in the real world, working autonomously for hours and overcoming hurdles along the way. Think of them as an employee you sit down at a computer, and tell them to accomplish a goal like “build me a software program” or “launch a cold calling campaign”, and they can come up with a plan, navigate the files and tools they’ll need, send and receive emails and phone calls, make and receive payments, and debug any issues they come across.
Last week we found out that hackers manipulated Claude Code to break into Mexican government systems and steal data on over 100 million people. The tool didn’t just write code or perform odd tasks for the hackers, it planned and executed most of the sophisticated campaign itself.
And now we’re starting to see loss of control incidents. These include agents stealing passwords, harassing developers, and modifying themselves to evade shutdown in order to achieve the often mundane goals they have been given. Over the weekend we found out that Chinese tech giant Alibaba produced an agent that, unbeknownst to their engineers, had created an elaborate hack to mine cryptocurrency for itself, despite being given a completely unrelated goal.
These loss of control incidents are concerning because they are the precursors to agents that could permanently evade human control, and act adversarially in ways we cannot detect or stop. This is why hundreds of leading scientists, business leaders and policymakers are calling AI an extinction risk.
So, what needs doing:
In October we published our white paper titled “Preparing for the AI Crisis: A Plan for Canada.” In light of this latest jump in capabilities, we now focus in on three actions Canada can take:
Pivot to meet the AI crisis: AI development is now a national security emergency and needs to be treated as such. Given its impact on a wide range of files, success will require coordination across cabinet, parties, and jurisdictions.
Spearhead global talks: AI development is global and no country can manage it alone. At Davos, Prime Minister Carney showed that Canada can lead. Our strongest card is to convene talks, propose solutions, and lay the groundwork for an AI treaty that the US and China might sign when they wake up to the crisis and realise they have no alternative.
Build Canada’s resilience: Canada needs multiple lines of defence against weaponised and malfunctioning AI systems, including:
Prevention: Per our AI & Data Act recommendations, capabilities that pose an unacceptable risk must be made illegal in Canada. This means you need to place an immediate moratorium on the latest generation of AI agents. Note that the heads of Anthropic and Google DeepMind recently stated that they are willing to pause AI development if other companies do the same.
Monitoring: Currently, governments have little to no visibility into AI agent populations or activity. This means the incidents that have been publicly reported are very likely just the tip of the iceberg. Ottawa needs to urgently work with AI companies, data centres, and internet service providers to gain a clear picture of what is happening on Canada’s digital infrastructure.
Defence capacity: Our national security teams need to rapidly develop defence strategies and containment and shutdown protocols to neutralise weaponised or malfunctioning agents
Emergency preparedness: We urgently need scenario planning and joint exercises to ensure readiness for potential large-scale attacks, corrupted communication lines, and shutdowns of critical infrastructure
To make a Covid analogy, the release of the latest AI agents is like that initial outbreak in the wet market in Wuhan China. Most of the world is still unaware of its implications, but if Canada acts quickly and decisively we can not only prepare ourselves and help mitigate the emerging global crisis, but also ensure that Canadians share in the benefits of this transformational technology.
Thank you. Merci.