The federal government has reconvened its advisory group on online safety as it considers legislating age restrictions for social media users, and measures to address the “emerging risks” posed by AI chatbots, The Logic has learned. 

On Thursday, the Department of Canadian Heritage requested input from the 12-member expert panel, which advises Canadian Heritage Minister Marc Miller on how to legislate harmful online content. A letter sent to members of the group said officials are considering “additional elements for a future legislative regulatory framework, such as exploring social media age restrictions and emerging risks surrounding AI.”

Bernie Farber, a member of the panel and the former chair of the Toronto-based non-profit Anti-Hate Network, shared parts of the letter with The Logic. 

“I am quite pleased with this,” Farber said. “The advisory group has already provided a road map of what a law should look like and how to deal with the new technology, but the government wants to hear more. I’m taking that as a positive sign.”

The reconvention of the panel marks a notable shift in Ottawa’s fast-changing response to potential dangers of web technology, social media and, increasingly, AI. The group formed in 2022 to help the government led by Justin Trudeau devise online harms legislation that never passed, despite two attempts by the Liberals to get it through Parliament. The panel was not, until now, a formal part of the government’s efforts under Prime Minister Mark Carney to address online harms.

This latest call for advice from the group comes nearly two weeks after 18-year-old Jesse Van Rootselaar killed eight people before killing herself in the community of roughly 2,400 people. The panel last met in June 2022. 

AI Minister Evan Solomon summoned OpenAI’s safety team to Ottawa this week after The Wall Street Journal reported that the San Francisco-based company behind ChatGPT knew about Van Rootselaar’s disturbing conversations with the chatbot, yet didn’t alert authorities. Unless OpenAI makes changes to its safety protocols, the government will tighten legislation governing the technology, Justice Minister Sean Fraser said after the meeting.

Miller’s press secretary Hermine Landry said the government has yet to make a decision on whether to legislate chatbots. “We all want our children to be safe as they navigate the digital world, and platforms have an important role to play in meeting that challenge. Our government intends to act swiftly to better protect Canadians, especially children, from online harm,” she said.

On Thursday, OpenAI vice-president of global policy Ann M. O’Leary sent a letter to Solomon listing several changes to its safety protocol, including establishing a direct point of contact with Canadian law enforcement and adapting its models to help de-escalate when users are in distress or pursuing prohibited behavior.

The legislative road to regulating online harms has been long. Under Trudeau, the government first proposed measures to curb hate and violence on social media platforms in 2021. The expert advisory group on online safety was formed in part to address concerns that these proposals would lead to the creation of a “surveillance state” by giving too much power to law enforcement and Big Tech.

The Liberals’ first legislative attempt to regulate online harms died on the order paper prior to the 2021 election. In February 2024, the government introduced the Online Harms Act, which focused on material that incites extremism, hatred or violence, as well as sexually exploitative content involving children. The bill, which contained no provisions regarding AI chatbots, also died on the order paper when Parliament dissolved in January 2025.

Now, under a new prime minister, the government is poised to try again. One senior official, speaking on background on the condition they not be named, told The Logic: “Online Harms has died twice before, so hopefully, three times a charm.”