The introduction of Bill C-34, the Safe Social Media Act (the “SSMA”) marks an ambitious attempt by the Canadian government to regulate social media, online services, and AI chatbots. The legislation would establish a new Digital Safety Act (DSA), create a Digital Safety Commission of Canada, impose safety obligations on social media and AI chatbot services, and prohibit social media accounts for children under 16 unless a service qualifies for an exemption based on child-safety safeguards.
Viewed in isolation, Bill C-34 may appear to be another online harms bill. It is not. The legislation represents a synthesis of several international regulatory trends that have emerged to protect the public from online, social media, and AI chatbot harms. It reflects key elements of the European Union’s Digital Services Act (“EU DSA”), Australia’s Online Safety Amendment (Social Media Minimum Age) Act 2024 social-media restrictions for children, and a growing body of U.S. state legislation aimed at AI companion chatbots.
Bill C-34’s purposes
Section 14 of the Bill sets out the purposes of the SSMA as follows:
(a) promote the safety of persons in Canada;
(b) protect children’s physical and mental health;
(c) considering that exposure to harmful content online impacts the safety and well-being of persons in Canada, mitigate the risk that persons in Canada will be exposed to harmful content online while respecting their freedom of expression;
(d) enable persons in Canada to participate fully in public discourse and exercise their freedom of expression online without being hindered by harmful content;
(e) enable persons in Canada to benefit from the use of chatbot services while reducing the risk of harm to the users of those services and other persons;
(f) reduce harms caused to persons in Canada as a result of harmful content online;
(g) make content that sexually victimizes a child or revictimizes a survivor and intimate content communicated without consent inaccessible online;
(h) ensure that operators are transparent and accountable with respect to their duties under this Act; and
(i) contribute to the development of standards with respect to online safety.
The bill does not create a general civil cause of action against platforms for online harms. Nor is it simply a notice-and-takedown statute. Its dominant structure is regulatory, most of which is unknown because it will need to be brought into effect via future regulations. It will create duties for operators of regulated services, empower a specialized Digital Safety Commission, require digital safety plans, enable compliance orders, and create an administrative monetary penalty regime with large AMPs of the greater of $10 million and 3% of gross global revenues.
The Act distinguishes three categories of services: regulated social media services, regulated chatbot services, and regulated online services within categories to be established by regulation. These services that will be covered if and to the extent designed by regulation to be a “covered” service are defined as follows:
social media service means a website or application that is accessible in Canada, the primary purpose of which is to facilitate interprovincial or international online communication among users of the website or application by enabling them to access and share content.
chatbot service means an artificial intelligence system that
(a) communicates over the Internet;
(b) is made available on a website or through an application that is publicly accessible in Canada;
(c) uses a natural language interface to provide, in a conversational format, adaptive, human-like responses to user inputs;
(d) is capable of being used, by means of multiple interactions or sessions, to simulate a sustained human-like relationship with a user, including one that may resemble friendship, an intimate relationship or therapeutic support; and
(e) generates content or responses that are not fully predetermined by the developer of the system or the person that operates the system.
online service means a website or application, other than a social media service or a chatbot service, that is accessible in Canada over the Internet and that allows users of the website or application to interact with the website or application.
Duties imposed on all regulated services
Each of the operators of the three regulated services has a duty to protect children by integrating into the service design features respecting the protection of children to be set out in regulations, implement age-verification or age-estimation measures that are adequate to mitigate the risk that children will be exposed to pornographic content on the service if they have reasonable grounds to suspect that the regulated service that they operate provides access to pornographic content, and implement any measures that are provided for by regulations to mitigate the risk that children who use the service will be exposed to pornographic content on the service.
Duties imposed on social media services
Operators of regulated social media services will have to comply with numerous new duties, most of which are yet to be established by regulation. In summary the obligations will include:
- Controversial requirements to implement an adequate age-verification or age-estimation measures designed to prevent a person under the age of 16 from being able to have an account with, or be otherwise registered with, the service. The obligation will apply only in respect of regulated social media services — or regulated social media services that fall within classes of regulated social media services — specified in regulations. The Commission may, on application by an operator, exempt, the operator from the requirements if the Commission is satisfied that the operator provides adequate safeguards in the regulated social media service for the protection of children. While this is intended to provide a nuanced approach to the social media ban, it is unclear how it will work and whether the process will enable exemptions to be in place before the ban takes effect.
- A duty to act responsibly in respect of a regulated social media service that it operates by complying with a detailed list set of requirements and with orders made by the Commission. The duties will include:
- implementing measures that are adequate to mitigate the risk that users of the service will be exposed to harmful content on the service;
- making available to users who have an account or are otherwise registered with the service tools that enable those users to block other users who have an account or are otherwise registered with the service from finding or communicating with them on the service;
- implementing any measures that are provided for by regulations in respect of the labelling of synthetic content or harmful content;
- submitting a digital safety plan to the Commission in respect of each regulated social media service that it operates;
- making certain content inaccessible to all persons in Canada; and
- investigating certain content flagged by users.
The labelling requirements cover a broad range of social media content because of the breadth of the definitions of synthetic and harmful content. These terms are defined as follows:
synthetic content means an audio representation or visual representation of a person, object, place, entity or event that is made by electronic or mechanical means, including by means of artificial intelligence software, if the depiction is likely to be mistaken for an authentic audio or visual recording of a person, object, place, entity or event.
harmful content means
(a) intimate content communicated without consent;
(b) content that sexually victimizes a child or revictimizes a survivor;
(c) content that induces a child to harm themselves;
(d) content used to bully a child;
(e) content that foments hatred;
(f) content that incites violence; and
(g) terrorism or violent extremism content.
The harmful content labelling requirements are even broader when taking into account how the terms within the definition of harmful content are defined. The definitions are set out below.
content that induces a child to harm themselves means content that advocates self-harm, disordered eating or dying by suicide or that counsels a person to commit or engage in any of those acts, and that, given the context in which it is communicated, could cause a child to inflict injury on themselves, to have an eating disorder or to die by suicide.
content that foments hatred means content that expresses detestation or vilification of an individual or group of individuals on the basis of a prohibited ground of discrimination, within the meaning of the Canadian Human Rights Act, and that, given the context in which it is communicated, is likely to foment detestation or vilification of an individual or group of individuals on the basis of such a prohibited ground
intimate content communicated without consent means
(a) a visual recording, such as a photographic, film or video recording, in which a person is nude, is nearly nude or is exposing their sexual organs or anal region or is engaged in explicit sexual activity, if it is reasonable to suspect that the person
(i) had a reasonable expectation of privacy at the time of the recording, and
(ii) does not consent to the recording being communicated; and
(b) a visual representation that is made by any electronic or mechanical means, including by means of artificial intelligence software, and that shows an identifiable person who is depicted as nude, as nearly nude, as exposing their sexual organs or anal region or as engaged in explicit sexual activity, if the depiction is likely to be mistaken for a visual recording of that person and if it is reasonable to suspect that the person does not consent to the representation being communicated.
content that sexually victimizes a child or revictimizes a survivor means
(a) a visual representation that shows a child, or a person depicted as being a child, who is engaged in or depicted as being engaged in explicit sexual activity;
(b) a visual representation that depicts the sexual organs or anal region of a child, if it is reasonable to suspect that the representation is created or communicated for a sexual purpose;
(c) written material or an audio recording whose dominant characteristic is the description, presentation or representation of explicit sexual activity with a child, if it is reasonable to suspect that the material or recording is created or communicated for a sexual purpose;
(d) a visual representation, written material or an audio recording that shows, describes, presents or represents any of the following, if it is reasonable to suspect that the representation, material or recording is created or communicated for a sexual purpose:
(i) a person touching, in a sexual manner, directly or indirectly, with a part of their body or with an object, any part of the body of a child or a person depicted as being a child,
(ii) a person who is engaged in or depicted as being engaged in explicit sexual activity in the presence of a child or a person depicted as being a child, or
(iii) a person exposing their sexual organs or anal region in the presence of a child or a person depicted as being a child;
(e) a visual representation, written material or an audio recording in which or by means of which sexual activity between a person who is 18 years of age or more and a child is advocated, counselled or planned, other than one in which or by means of which sexual activity between a person who is 16 years of age or more but under 18 years of age and another person who is less than two years older than that person is advocated, counselled or planned;
(f) a visual representation that shows a child who is being subjected to cruel, inhuman or degrading acts of physical violence, whether the acts are sexual in nature or not;
(g) any excerpt of a visual representation referred to in paragraph (a), if it is reasonable to suspect that the communication of the excerpt perpetuates harm against a person who as a child appeared in the visual representation; and
(h) a visual representation, written material or an audio recording that, given the context in which it is communicated, is likely to bring to light a connection between a person and a visual representation, written material or audio recording referred to in any of paragraphs (a) to (d) in which the person appeared as a child, if it is reasonable to suspect that the communication of the representation, material or recording that is likely to bring to light that connection perpetuates harm against the person.
content used to bully a child means content, or an aggregate of content, that, given the context in which it is communicated, could cause serious harm to a child’s physical or mental health, if it is reasonable to suspect that the content or the aggregate of content is communicated for the purpose of threatening, intimidating or humiliating the child.
terrorism or violent extremism content means content that, given the context in which it is communicated,
(a) actively encourages, instructs or counsels a person to facilitate or commit an act of terrorism or violent extremism;
(b) threatens or facilitates the commission of an act of terrorism or violent extremism;
(c) aims to recruit a person to facilitate or commit acts of terrorism or violent extremism;
(d) offers or provides training, skills or expertise to a person so that they or another person may facilitate or commit an act of terrorism or violent extremism;
(e) actively encourages a person to collect, provide or make available a material benefit, including a financial benefit, to facilitate or commit an act of terrorism or violent extremism; or
(f) promotes an act of terrorism or violent extremism in a manner that could reasonably be expected to encourage others to facilitate or commit acts of terrorism or violent extremism.
The chatbot provisions
A novel aspect of Bill C-34 is its direct regulation of AI chatbot services. The bill would impose a duty to act responsibly on operators of regulated chatbot services. That duty includes taking measures adequate to mitigate the risk that the chatbot will communicate harmful content to a user. This obligation is very broad considering the definition of “harmful content”. It also includes crisis-intervention duties and duties to mitigate harmful behavior by the chatbot itself.
The crisis-intervention provision is especially important. If a user expresses suicidal ideation, an intention to self-harm, or an intention to commit an act that could cause death or serious bodily harm to another individual, the operator must implement measures that cause the service to immediately interrupt its interaction with the user in order to direct the user toward crisis intervention services that are appropriate, available at the moment, and permit interaction with a human being.
The bill also identifies types of chatbot “harmful behaviour” that operators must mitigate. These include posing as a human being in a manner likely to lead a user to mistake it for a human or otherwise being deceptive about being AI; posing as a medical, legal, or other licensed professional and giving advice based on that deception that could reasonably be expected to be relied upon; using manipulative engagement techniques to encourage emotional attachment in a way that may encourage social withdrawal or disconnection from reality; encouraging self-harm, suicide, or acts that could cause death or serious bodily harm; and any other behaviour later specified by regulation.
Online services regulation
The DSA would also regulate online services, being very broadly defined to be a website or application, other than a social media service or a chatbot service, that is accessible in Canada over the Internet and that allows users of the website or application to interact with the website or application.
In addition to obligations being imposed on all three regulated services, the specific obligations of regulated online services consists a requirement to submit a digital safety plan to the Commission in respect of each regulated online service that it operates.
Influences of other online harms laws
Bill C-34 has been influenced by three major sources.
EU DSA
One influential source of the DSA is the EU DSA. The EU DSA represented a major shift in European regulation because it moved beyond a notice-and-takedown model and imposed broader obligations relating to systemic risks, platform design, transparency, and the protection of minors.
Bill C-34 appears to follow a similar basic philosophy. Both regimes focus on preventing harms before they occur rather than merely responding after the fact. Both emphasize platform accountability, risk management, and safety-by-design principles. Both also place particular emphasis on protecting children.
While Bill C-34 uses different legislative machinery, it is animated by many of the same concerns. The Canadian proposal similarly seeks to address harms arising from platform architecture and business models rather than focusing solely on individual pieces of content.
In that sense, Bill C-34 can be viewed as part of a broader international movement toward regulating the design and operation of digital services.
Australia’s Influence: Child protection through access controls
Australia provides another important point of comparison.
Australia’s Online Safety Amendment (Social Media Minimum Age) Act 2024 requires covered social-media services to take reasonable steps to prevent users below a specified age threshold from holding accounts.
Bill C-34 adopts a similar premise: that social-media services can create significant risks for children and that those risks justify age-based restrictions and platform obligations.
There are, however, important differences. Australia’s legislation focuses primarily on access restrictions. Bill C-34 is considerably broader. It combines age restrictions with platform duties, chatbot regulation, a new regulator, administrative enforcement powers, and a broader digital-safety framework.
Accordingly, it may be more accurate to describe Bill C-34 as adopting Australia’s child-protection objectives while pursuing them through a more comprehensive regulatory architecture.
The U.S. Model: regulating AI companion chatbots
The most novel aspect of Bill C-34 is its treatment of AI chatbots.
As I noted in my testimony to the INDU Committee recently, some U.S. states have enacted laws to regulate the uses of chatbots and AI systems that can promote self harm. These include states such as Main, New York which has several laws, California, Washington, and Wyoming. The common themes in these state laws are:
- transparency;
- disclosure that users are interacting with AI;
- child protection;
- intervention mechanisms relating to self-harm;
- safeguards against emotional dependency; and
- accountability for foreseeable chatbot-related harms.
Bill C-34 appears to reflect many of the same concerns. Unlike some of the state laws, however, Bill C-34 does not focus exclusively on companion chatbots. Instead, it incorporates chatbot regulation into a broader digital-safety framework.
Bill C-34 also does include some of the other regulatory requirements for high risk AI systems that have been enacted in U.S. states such as transparency measures for frontier models such as those enacted in California and NewYork, and safety measures for AI controlled critical infrastructure such as enacted in Montanna. Importantly, and to the credit of the government, the DSA eschews the flawed approach that would have been adopted in AIDA or the heavily criticized EU AI Act.
A Different Regulatory Philosophy
Perhaps the most important observation is that Bill C-34 does not simply copy any existing foreign model. The EU DSA is primarily concerned with systemic platform risks. Australia’s legislation is principally concerned with restricting children’s access to social-media services. Some U.S. state laws focus on the unique risks associated with AI companion systems. Bill C-34 combines all three approaches.
- From Europe, it borrows platform accountability and systemic-risk concepts.
- From Australia, it borrows the idea that children require enhanced protection and that age-based restrictions may be appropriate.
- From the United States, it borrows concern about the distinctive risks created by conversational AI systems and emotionally responsive chatbots.
The result is a uniquely Canadian hybrid.
Key Questions Going Forward
While the objectives of the legislation are relatively clear, almost all of the details remain uncertain. Just like the flawed AI bill, AIDA, before it, the bill delegates the most key aspects of the law to regulations. Some may say this is the most flexible way to address fast moving technologies. Other my say this is another unfortunate delegation of parliamentary supremacy over key laws to unaccountable officials.
Whether Parliament ultimately adopts the bill in its current form remains uncertain. What is already clear, however, is that Canada has joined a growing group of jurisdictions that are moving beyond content moderation and toward a broader model of platform accountability, child protection, and certain forms of AI safety.
