New legislation would bring AI chatbot services into Canada’s online harms framework, creating compliance obligations.
The Government of Canada introduced legislation that would extend new online safety requirements to AI chatbot providers, bringing AI services into a broader effort to reduce online harms affecting children.
Bill C-34, the Online Harms Act, would require social media platforms and AI chatbot services to meet new online safety requirements. While public attention has largely focused on proposed restrictions affecting youth access to social media, the legislation also establishes new expectations for organizations operating AI chatbots in Canada.
The legislation would create a new Digital Safety Commission to oversee regulated services and enforce compliance. A Digital Safety Ombudsperson would support users and help navigate complaints, while a Digital Safety Office would provide administrative support.
Under the proposal, regulated services would be required to take measures to reduce users’ exposure to harmful content and activity. The government said the legislation is intended to address harms including content that sexually victimizes children, promotes self-harm, incites violence, or facilitates the exploitation of children online.
AI chatbot services are specifically included within the scope of the government’s child-safety initiative. According to the Department of Canadian Heritage, the legislation is designed to ensure that both social media services and AI chatbots implement safeguards that better protect young users.
The legislation does not yet specify all obligations that would apply to AI chatbot providers. While the bill would bring AI chatbot services within the scope of the online harms framework and under the oversight of the proposed Digital Safety Commission, the precise compliance requirements and their impact on developers would depend on the final text of the legislation and any accompanying regulations.
Canada is the latest country to propose legislation around AI and child safety. The European Union recently finalized the EU AI Act, banning, among other things, AI-generated child sexual abuse material (CSAM). In the United States, individual states such as California and Illinois passed laws that protect minors from AI chatbots.
If adopted, the legislation would establish one of Canada’s most significant regulatory frameworks governing AI chatbot safety and child-protection obligations.
