
Joseph Kostousov isn’t worried about AI chat bots. Instead, the second-year Economics Specialist is investigating the weaponization of autonomous artificial intelligence, and other catastrophic risks associated with loss-of-control. He founded the Toronto AI Safety Initiative (TAISI) to investigate the issue further at the start of the 2025-26 year.
“I learned about the AI safety problem a bit in my last year of high school, and I thought, this is pretty bad. This is coming a lot sooner than we think,” the St. Micheal’s College student explained. “The future could be a cool place if we get this right, but if we don’t, some AI experts, including Geoffrey Hinton, are predicting a 10 to 20% chance of human extinction within the next three decades. It just felt like an urgent issue, and it’s kind of neglected.”









