Canada is collaborating with the U.K. on a new artificial intelligence safety initiative as the U.S. pushes its vision of a zero-sum AI race that will see it remove guardrails on development and deployment of the burgeoning technology.

On Wednesday, Canada announced that it will commit $1 million to a $29 million joint AI safety effort with the U.K. that will bankroll research and commercial projects focused on keeping advanced AI systems in line — in other words, ensuring that they operate safely, reliably, and in a useful way, and without unintended or harmful actions.

“Together, we’re advancing cutting-edge research to ensure (the) next generation of AI systems are not only powerful but also reliable — serving societies here at home and around the world,” said minister of AI and digital innovation Evan Solomon in a statement.

The initiative, called the AI Alignment Project, is spearheaded by the U.K.’s AI Security Institute (AISI) and will involve the Canadian AI Safety Institute (CAISI) under the umbrella of the Canadian Institute for Advanced Research (CIFAR), a research organization focussed on innovation and deep tech. Other financial backers include enterprise partners such as Amazon Web Services, Inc. and venture firm the Safe AI Fund, and non-profits such as New-York headquartered Schmidt Sciences, LLC.

The project will provide grant funding for researchers across disciplines including computer and cognitive science; help organizations and individuals access venture capital investment; and secure “compute,” the computational power needed to train and run AI models. The call for proposals, which launches Wednesday, will be open until September.

Prominent AI experts including Canadian-French computer scientist

Yoshua Bengio , known as one of the “godfathers of AI,” will serve on the advisory board to help steer the effort and to select the successful proposals by November.

Canada’s participation in the U.K.-led AI safety effort fits into Ottawa’s broader AI vision for Canada — one that involves building trust in, and adoption of, the technology — and lets Canadian researchers have a seat at the table when it comes to global efforts on AI safety, according to Elissa Strome, executive director of pan-Canadian AI strategy at CIFAR. There are “few challenges more urgent than ensuring AI is safe, predictable, and beneficial for all,” she said.

The launch of the new initiative takes place amid broader shifts in the conversation on AI. As the tech becomes an increasingly important feature of national security and economic competition, the U.S. has peddled a light-touch regulatory approach, while jurisdictions such as the EU have championed tougher-on-tech rules. Industry groups and AI safety scholars have also clashed over how AI should be regulated, developed and deployed.

Ottawa so far, has failed to pass comprehensive AI regulation, and the Carney government has so far, divulged few details on how they might regulate the technology.

Bill C-27 — legislation that focused on modernizing Canadian data privacy and protection rules, in addition to the Artificial Intelligence and Data Act (AIDA) — died in January when the former Trudeau government prorogued parliament.

Strome said: “They’re taking this time to think very carefully about what direction Canada will go. They’re taking stock, talking to stakeholders, and watching closely what’s happening in other jurisdictions … and trying to put together an approach that will work.”

  • What does Trump’s ‘free-for-all’ AI action plan mean for Canada?
  • What are AI agents and which jobs are they coming for first?

• Email: ylau@postmedia.com Bookmark our website and support our journalism: Don’t miss the business news you need to know — add financialpost.com to your bookmarks and sign up for our newsletters here.