As worries grow about instances of AI agents going rogue, experts and stakeholders are increasingly advocating for government regulation of AI systems that are becoming more advanced. However, some experts argue that global cooperation and the rapid pace of AI development could pose challenges to such regulation efforts.
While individuals like U.S. President Donald Trump express confidence in tech giants’ ability to self-regulate, other industry leaders such as Microsoft co-founder Bill Gates and Canadian-British AI pioneer Geoffrey Hinton assert that AI necessitates safeguards beyond self-regulation. Jaxson Khan, an expert in AI policy and adoption at the University of Toronto’s Munk School of Global Affairs, emphasized the need for government oversight in AI, similar to other regulated industries like food or aviation.
Acknowledging the swift evolution of AI models and their increasing capabilities, experts highlight the difficulty for governments to keep pace with these rapid changes. The exponential growth in AI agents’ capabilities every four months, as reported by Anthropic, and the potential for AI models to recursively self-improve pose unique challenges for regulation.
Calls for global oversight led by a non-industry body to manage and mitigate risks associated with advanced AI systems have been made by experts like Prime Minister Mark Carney and AI pioneers Geoffrey Hinton and Yoshua Bengio. While challenges in reaching international consensus on AI governance persist, establishing such a body could facilitate collaboration among countries to enhance system resilience and prevent AI misuse.
In Canada, there is currently no comprehensive federal regulation of AI. However, proposed legislation such as Bill C-34, the Safe Social Media Act, sets standards for AI chatbots to minimize the generation of harmful content. Initiatives like the national council on artificial intelligence, introduced by Carney, aim to guide the government in formulating an AI strategy and ensuring technology safety.
Although limitations exist in the federal government’s ability to regulate multinational AI companies, policies requiring incident reporting and accountability could be implemented. By promoting transparency and holding companies responsible for incidents affecting Canadians, the government can take steps towards educating citizens and demonstrating leadership in AI governance.