Federal ministers voiced disappointment Wednesday after OpenAI representatives failed to present concrete steps to improve AI safety, particularly concerning the notification of law enforcement regarding user behavior. This meeting followed concerns raised after the Tumbler Ridge, B.C., mass shooting, prompting questions about the government's slow approach to artificial intelligence regulation.
Artificial Intelligence Minister Evan Solomon stated that while legislation addressing AI safety and privacy is in development, he is awaiting "concrete" proposals from OpenAI before proceeding. The company was summoned to Ottawa after it was revealed the Tumbler Ridge shooter had been flagged internally for activity on the AI chatbot months prior, but OpenAI did not alert the RCMP until after the shooting occurred.
Justice Minister Sean Fraser and Public Safety Minister Gary Anandasangaree were also present at the meeting. Prime Minister Mark Carney indicated a willingness to consider changes that could prevent future tragedies. Ministers emphasized that any government action would focus on the threshold for escalating concerning behavior to law enforcement.
OpenAI stated they appreciated the "frank discussion" and are committed to following up with an update on additional steps being taken. However, experts argue that the AI industry should not be left to self-regulate and that government intervention is necessary. Efforts to regulate AI and address online harms through legislation have faced delays in Parliament.
While allies like the UK and EU are strengthening AI regulations, the U.S. approach has been sporadic. Canada's current privacy legislation allows companies discretion in disclosing personal information to authorities, leading to internal thresholds like OpenAI's "imminent" threat assessment. The government is working to update relevant privacy legislation.