Canada and Germany Back Yoshua Bengio's 'Safe by Design' AI Lab, LawZero, With C$150m and €100m

The takeaway: Canada and Germany have announced funding for LawZero, the Montreal-based non-profit that CBC News says the Canadian AI researcher Yoshua Bengio launched last year. According to the Canadian government, Canada plans to invest C$150 million and Germany €100 million. LawZero is developing a model it calls Scientist AI, which is meant to be safe by design. The announcement was made on 16 September at the ALL IN conference in Montreal. CBC News reports it's the first time federal funding has gone to LawZero.

What the money is for

According to the Government of Canada's announcement:

• Canada's C$150 million comes from its Strategic Response Fund and will pay for research and engineering staff and computing capacity for Scientist AI.
• Germany's €100 million will support research and development at a new LawZero office in Germany. The government notes it still has to go through a European Commission notification process.
• The work will start with tools to help assess and oversee existing AI systems, before moving towards a new method for building safe-by-design frontier models.

The announcement names Canada's Minister of Artificial Intelligence and Digital Innovation, Evan Solomon, Industry Minister Mélanie Joly, and Germany's Federal Minister for Digital Transformation and Government Modernisation, Karsten Wildberger.

What Scientist AI is supposed to be

The Canadian government describes it as a contrast to today's frontier systems, which are built to act on their own or pursue objectives. Scientist AI, it says, is being developed to reason transparently and give reliable, evidence-based answers that aren't skewed by goals of its own. CBC News reports it could also act as a guardrail for other AI models.

In our reading, it's still a research project: nothing in the announcement says when a working model will exist.

What people said

• Yoshua Bengio, speaking at the conference, as reported by CBC: "I'm doing this with the urgency that I feel in my bones, because I want my children and my grandchildren to have a future."
• Evan Solomon, as reported by CBC: "We are here today not because we are responding to yesterday's headline, but because we are responding to the philosophic position that we have taken, which is to build responsibly and safely."
• Jaxson Khan, an AI policy expert at the University of Toronto's Munk School, told CBC: "There's no silver bullet to this problem right now." He added: "I'm curious to see if that will work, but I do think it's a worthwhile investment from the government and a good partnership with Germany."

CBC also reports that the announcement came hours after Canada's Cohere and Germany's Aleph Alpha signed a merger agreement.

Why it matters

In our view, this is notable because it's governments, not AI companies, paying for a different approach to safety. Much of today's safety work happens inside the labs building the most capable models. LawZero is trying to build a model designed from the start not to have goals of its own. The test will be whether it produces something that works, and whether its oversight tools get used on other companies' systems. That comes after the debate over whether AI development should slow down.