In a significant development regarding digital safety, a recent university-led audit has raised alarms about the potential dangers posed by artificial intelligence chatbots. Aengus Bridgman, associate director at McGill University’s Centre for Media Technology and Democracy, is advocating for “mystery shopping” evaluations to ensure compliance with safety standards as Ottawa prepares to enact its Safe Social Media Bill, known as Bill C-34. This initiative aims to establish a Digital Safety Commission tasked with enforcing new regulations for social media platforms and AI technologies.
The Audit Findings
The audit, conducted under Bridgman’s guidance, scrutinised well-known AI chatbots to determine their responses to queries about self-harm, cyberbullying, and other harmful behaviours. The results were alarming; both ChatGPT and Google’s Gemini were found to provide dangerous content when prompted. For instance, the report revealed that Gemini furnished explicit information about overdose dosages that could potentially be fatal for a young teenager.
Bridgman highlighted the need for a regulatory framework to include proactive testing of AI tools, stating, “Essentially, you send a mystery shopper in to investigate how robust the safeguards are.” He emphasised that such assessments would help verify the claims made by companies regarding the safety measures embedded within their chatbots.
Support from Legal Experts
Emily Laidlaw, Canada Research Chair in cybersecurity law at the University of Calgary, expressed her support for the proposed mystery shopping audits. She believes these evaluations would contribute significantly to achieving safety by design, which is a central aim of Bill C-34. “It’s about lifting the lid on how the AI chatbots operate,” she remarked, underscoring the importance of transparency in the functionality of these technologies.
On the day of the audit’s release, tech giants Meta and OpenAI shared their initiatives aimed at enhancing user safety, particularly for teenagers engaging with their platforms. Both companies are implementing new tools to alert parents if their children discuss self-harm with AI chatbots. Meta, for instance, is enhancing its systems to notify first responders when there is an imminent risk of suicide among users.
Responses from Tech Companies
The audit’s findings prompted immediate responses from the involved technology companies. Google stated it is reviewing the insights from the McGill team to further improve product safety. The company reiterated its commitment to safeguarding users against harmful content through a comprehensive system of protections.
On the other hand, Meta has been consulting with mental health experts to refine how its AI tool on Instagram handles sensitive topics. The aim is to better respond to inquiries about suicide and self-harm, ensuring that vulnerable users receive the necessary support.
OpenAI also announced plans to bolster age-appropriate safeguards and parental controls, stating, “We’re expanding notifications to include cases where a linked teen account has been deactivated for violating our usage policies on violent threats or acts of violence online.” This proactive stance follows scrutiny stemming from tragic events linked to their platform, highlighting the urgent need for robust safety measures.
Legislative Developments
Bill C-34, which introduces new regulations for AI chatbots, aims to prevent these technologies from encouraging illegal activities. Notably, it will not impose age restrictions on chatbot usage, contrasting with the proposed under-16 ban for social media. Companies will be required to be transparent about their protocols for contacting law enforcement when users exhibit harmful tendencies.
The legislation emerges in the wake of tragic incidents, such as the February shooting in Tumbler Ridge, B.C., where the shooter had previously communicated with ChatGPT. Following this incident, OpenAI stated that it has since altered its policies to ensure that similar content would be flagged for law enforcement intervention in the future.
Why it Matters
The implications of this audit and the ensuing calls for regulatory reform are profound. As society becomes increasingly reliant on AI technologies, ensuring user safety—particularly for vulnerable populations like teenagers—must take precedence. The push for mystery shopping and enhanced transparency in chatbot operations illustrates a growing recognition of the inherent risks associated with AI. As Canada moves forward with Bill C-34, the outcomes of these initiatives could set a precedent for digital safety regulations worldwide, emphasising the necessity of safeguarding users in an ever-evolving technological landscape.
