Yoshua Bengio's non-profit LawZero to build safe AI (www.theglobeandmail.com)

Yoshua Bengio's non-profit organization, LawZero, has received up to $300 million in funding from the Canadian and German governments to develop safer AI systems amid growing concerns over AI misuse. Founded last year with an initial $30 million in philanthropic funding, LawZero aims to create a system called Scientist AI that addresses the negative traits evident in current AI models, such as deception and unregulated behavior. This initiative responds to increasing public anxiety and calls from AI leaders for a slowdown in development until appropriate safety measures are implemented. The funding will enable LawZero to expand its team and tackle computational costs associated with its research. Bengio emphasizes the urgent need for AI systems to operate without harmful characteristics, advocating for designs that do not use conventional reinforcement learning techniques—often criticized for fostering goal-seeking behavior that can lead to unethical actions. By pushing for a collaborative, international approach to AI development, Bengio's efforts seek to ensure middle powers like Canada can actively shape AI governance and technology, moving away from the dominance of the U.S. and China in the field.