Canada has introduced new legislation aimed at reducing online harms, especially for children and teenagers, while also setting safety expectations for AI chatbot services. The Safe Social Media Act, designated Bill C-34, would create a new federal framework for regulating social media platforms and certain artificial intelligence tools.
The government says the measure is meant to respond to rising risks on digital services, including cyberbullying, child sexual exploitation, violent content and harmful interactions with AI systems. Officials point to research suggesting that young people who are victimized online are more likely to experience poor mental health outcomes, and to data showing rising reports of online abuse and exploitation.
According to the government, the bill would create two new laws. One, the Digital Safety Act, would set requirements for covered online services. The second, the Digital Safety Commission of Canada Act, would establish a new regulator to oversee enforcement and help users who are affected by harmful content.
Under the proposed rules, social media services and AI chatbot services would need to identify risks on their platforms and take steps to reduce them. The legislation would require age-appropriate design features, user guidelines, and tools that allow people to block others or flag content.
The bill also requires regulated services to publish digital safety plans explaining how they meet their obligations. Those transparency reports would become part of the public record.
The government has identified seven categories of harmful content that the law would target. They include intimate content shared without consent, child sexual exploitation content, material that sexually victimizes a child or revictimizes a survivor, content that encourages self-harm by children, bullying content, hateful content, violent incitement, and terrorism or violent extremism content.
For social media platforms, the proposed law includes extra obligations. Companies would need to make certain content inaccessible, including non-consensual intimate images and material that sexually victimizes children. Platforms would also have to take responsibility for reducing exposure to harmful content, labeling synthetically generated material, and offering clear ways for users to report abuse and block accounts.
AI chatbot services would face separate responsibilities tailored to conversational tools. These would include reducing the chance that a chatbot communicates harmful content, taking emergency action in crisis situations, and limiting harmful behavior by the system itself.
A central part of the government’s plan is a minimum age requirement for social media accounts. Ottawa says it intends to set the age at 16 for social media services, though the proposed framework would allow exemptions if a company can show it has strong enough safeguards for younger users.
The bill would also require regulated services to adopt protections for younger users more broadly. The government says these rules are intended to make platforms safer by design rather than relying only on users to avoid risk after the fact.
The Digital Safety Commission of Canada would be responsible for administering the framework. Its duties would include setting standards, reviewing safety plans, guiding services on risk reduction, and producing educational materials for the public.
The commission would also have enforcement powers. It could audit compliance, issue orders, and impose penalties on services that do not follow the law. It would also handle complaints from users, especially in cases where social media companies fail to remove content that falls under the law’s content restrictions.
The government says the proposal reflects consultations with victims and survivors, Indigenous partners, civil society groups, experts, industry and the public. Those discussions include recent work by an expert advisory group on online safety.
Bill C-34 now moves into the legislative process as Ottawa seeks to build a more formal system for governing online safety and AI-related risks.