Canada proposes social media ban for under-16s with tech firm loopholes

Canada is moving toward a social media ban for anyone under 16, but with a crucial difference from the approach taken elsewhere: technology companies can escape the restriction entirely if they can show they have real policies in place to protect young users from harm. The Safe Social Media Act, introduced in the House of Commons on Wednesday by Culture Minister Marc Miller, represents the government's latest attempt to address what has become an urgent political question—how to keep children safe online while the companies that profit from their attention resist meaningful oversight.

The proposal arrives in the shadow of a tragedy. In February, an 18-year-old gunman killed eight people, including six children, in a school shooting in British Columbia. Before the attack, he had used ChatGPT to discuss gun violence. OpenAI, the company behind the chatbot, did not report his account to authorities. CEO Sam Altman later apologized in writing to the families of the victims. That failure—a company's decision not to flag a user discussing violence to law enforcement—has crystallized the stakes of the debate for Canadian policymakers.

Miller framed the legislation bluntly. "Kids are dying," he told reporters earlier in the week. "We will take all reasonable measures to make sure kids are safe in this country." The law would establish a new independent regulator, the Digital Safety Commission of Canada, whose members would be appointed by the cabinet. It would also impose sweeping rules on AI chatbots and what the government calls "harmful content" online. Companies that violate the rules face penalties of up to $10 million or 3 percent of their global gross revenue, whichever is larger.

But the exemption clause is the law's defining feature—and its potential weakness. Unlike Australia, which passed a strict ban last year that simply prohibits anyone under 16 from creating new accounts on platforms like Instagram, Facebook, and TikTok, Canada's version offers tech firms a way out. If a company can demonstrate it has adequate safeguards, it does not have to enforce the age restriction. Sara Austin, whose organization Children First Canada has long pushed for online safety legislation, sees this as a positive. The exemption, she argues, creates an incentive for companies to adopt better protections across the board, benefiting not just children but all users.

Yet Australia's experience suggests caution. That country's ban took effect with high expectations. But in a government survey, roughly 70 percent of parents reported their children were still using social media afterward. Many said platforms never asked them to verify their age. The Australian government has since opened five investigations into alleged violations by Facebook, Instagram, Snapchat, and TikTok. The question now is whether Canada's softer approach—one that trusts companies to police themselves if they commit to doing so—will fare better.

Canada has been slow to act on this front. The previous Liberal government tried twice to pass online safety legislation and failed both times. Other countries have moved faster. The UK has its Online Safety Act. France and New Zealand have passed their own measures. Greece is set to implement a ban for children under 15 in January. The timing of Canada's proposal is deliberate: world leaders gather for the G7 summit in France next week, where AI and child protection online are expected to dominate the conversation.

Free speech advocates have raised concerns that the law could expand censorship, though the government has not clarified exactly what criteria the new regulator will use to define harmful content or how it will enforce them. When the BBC asked the justice and culture ministries for details, neither responded. That silence itself speaks to the law's unfinished state—it is a framework without all its walls built, a set of intentions without complete specifications. What happens next depends on whether the Digital Safety Commission, once appointed, can actually compel compliance, and whether tech companies will treat the exemption as a genuine incentive to change behavior or simply as a loophole to exploit.