CMTD releases new research on social media safety, age verification, and AI governance amid major shift in Canada’s digital policy landscape
June 26, 2026 - As Canada advances a series of major digital policy initiatives, including the National AI Strategy, the Safe Social Media Act (Bill C-34), and Bill C-36, the Protecting Privacy and Consumer Data Act (PPCDA), the Centre for Media, Technology and Democracy and the Media Ecosystem Observatory are releasing a new body of research examining some of the most consequential questions facing policymakers: how to protect young people online, how age assurance systems should be governed, and how AI chatbot products cause harm and could be regulated within emerging online safety frameworks.
The four publications combine original public opinion research, comparative policy analysis, and technical audits of AI products. Together, they provide evidence to inform parliamentary debate and public discussion at a moment when Canada is considering significant reforms to the governance of online platforms and AI systems.
“With the tabling of C-36, Canada now has a new framework for social media and chatbot regulation. Now comes the hard part - evaluating, debating, and amending the proposed approach, and then if this bill passes implementing the regulations. This will demand ongoing research into how these digital products are shaping the lives of Canadians, and serious debate about policy design. We will continue to strive to do this work.”
Taylor Owen, Founding Director of CMTD and co-author on the AI audit briefs
Among the findings:
A national survey of 1,306 Canadians found broad support for stronger protections for young users online, particularly measures that place responsibility on platforms through age-appropriate design requirements, stronger legal accountability, and independent oversight.
A comparative analysis of age verification systems concludes that age assurance technologies are technically feasible and already deployed at scale internationally, but that their effectiveness depends on governance, privacy protections, and regulatory oversight rather than technology alone.
The first systematic audit of major AI chatbot products against harms contemplated in Canadian legislation found substantial variation in safety performance across providers. While one system refused harmful requests in 98 per cent of tests, another generated harmful content in nearly two-thirds. The AI findings carry particular significance for ongoing policy discussions. The audit found that safety outcomes depend not only on the capabilities of the underlying AI model, but also on the safeguards, product design choices, and governance measures implemented by providers. This distinction has important implications for how Parliament defines the scope of future AI and online safety regulation.
Rather than identifying a single technological solution, the research points to a common conclusion across both social media and AI: public safety outcomes are shaped by the governance choices made by companies and the accountability frameworks established by regulators.
Publications released this month include:
Online Harms AI Audit: Technical Brief & Online Harms AI Audit: Policy Brief (released today)
Age Verification for Online Platforms: Policy Considerations for Canada with Reference to the Adult Content Sector (released last week)
Survey on Canadians’ Preference for Social Media Age Verification Policies (released last week)
The Centre for Media, Technology and Democracy (CMTD) at McGill University produces independent research on the governance of digital platforms, information integrity, and the future of democracy. The Media Ecosystem Observatory (MEO) is a CMTD initiative tracking Canada's information ecosystem across platforms.
Lisez le communiqué de presse français.
Media Contact:
Isabelle Corriveau
Associate Director, Public Engagement
media@mediatechdemocracy.com
