Canada’s Bill C-34 asks social media and AI companies to take responsibility for risks their services create. A lawsuit arising from the Tumbler Ridge tragedy shows why the details matter

By David Bruno  |  Opinion  |  September 2026

THUNDER BAY – Technology – When a social media company recommends a harmful post, or an AI chatbot recognizes a dangerous exchange, responsibility cannot rest entirely with the person looking at the screen. The company designed the system, chose its safeguards and decided when to escalate a warning.

I made that argument in my earlier writing on online harms. Canada’s proposed Bill C-34, the Safe Social Media Act, offers a chance to put it into practice. It would require regulated services to reduce specified risks, protect children and publish plans explaining what safety measures they use and whether those measures work. A new Digital Safety Commission would oversee the rules. The bill is still before Parliament; its protections are not yet law.

The stakes became painfully clear after the February shooting in Tumbler Ridge, British Columbia. On September 21, the province filed a lawsuit against OpenAI in California, alleging that the company should have alerted law enforcement about warning signs in a user’s ChatGPT conversations before the attack. Families and survivors have also brought claims. These are allegations before the courts; the lawsuits have not established legal responsibility.

OpenAI has acknowledged that it banned an account associated with the attacker in June 2025 after automated detection and human review. The company says the information then available did not meet its threshold for a police referral. In a letter to federal ministers, it said its revised protocol would refer that same account if it were discovered today. It also said it later learned of a second account and shared it with law enforcement.

That sequence raises a public question beyond any one company: Who sets the threshold for action, who checks it, and what happens when a warning falls just below it? A blanket rule to report every disturbing conversation would threaten privacy and overwhelm police. A private threshold that nobody can examine leaves the public unable to tell whether a safety system is working.

Bill C-34 begins to answer that question. It proposes duties for AI chatbots to mitigate harmful outputs and behaviour, including encouraging violence, and to interrupt certain crisis conversations with a direction to available human help. Social media services would have to reduce exposure to defined harmful content. Safety plans would disclose risks, measures and evidence of effectiveness to the regulator and the public.

But a plan is only as good as the evidence behind it. Parliament should require companies to test their safeguards, document serious failures, and report whether changes actually reduce exposure and harm. The regulator needs the power and expertise to inspect that evidence. Users need a fair way to challenge decisions, while privacy and freedom of expression remain protected.

The B.C. case will be decided on its facts and the law that applies to it. It should not be used as a shortcut to assume that a chatbot caused the attack or that one report would certainly have prevented it. It does, however, show the cost of leaving vital safety decisions inside systems the public cannot scrutinize.

Canadians deserve innovation that can answer a simple question: When your service detects a serious risk, what did you do, and how do you know it was enough? That is the standard Bill C-34 should make real.

David Bruno is a cybersecurity expert and the author of a 2023 white paper on online harms.

Source notes

Parliament of Canada, Bill C-34, first reading: https://www.parl.ca/DocumentViewer/en/45-1/bill/C-34/first-reading

Parliament of Canada, LEGISinfo status: https://www.parl.ca/legisinfo/en/bill/45-1/c-34

Government of British Columbia, Attorney General statement, September 21, 2026: https://news.gov.bc.ca/releases/2026AG0067-001105

OpenAI, letter to Minister Evan Solomon, February 26, 2026: https://cdn.openai.com/pdf/8e938d69-0b67-4994-b9ff-683733ed587e/openai-letter-minister-solomon.pdf