**
In a significant push to ensure the safety of digital interactions, Ottawa is being urged to implement “mystery shopping” evaluations of artificial intelligence chatbots. This call to action follows a recent audit, which revealed alarming tendencies among popular AI platforms to inadvertently provide harmful advice regarding self-harm and cyberbullying. The recommendations come in light of federal Bill C-34, which aims to establish a Digital Safety Commission tasked with enforcing new safety standards within the burgeoning realm of social media and AI chatbots.
The Audit Findings
Aengus Bridgman, the associate director of the Centre for Media Technology and Democracy at McGill University, played a crucial role in the audit, which specifically targeted how AI chatbots respond to sensitive inquiries. The research, released in late June, aimed to scrutinise whether these tools adhere to safety protocols designed to protect vulnerable users. Bridgman emphasised that conducting regular assessments to check the efficacy of safety measures embedded in chatbots should be a cornerstone of the regulatory framework put forth by the federal government.
“Essentially, you send a mystery shopper in to investigate how robust the safeguards are,” Bridgman stated, highlighting the need for a proactive approach in monitoring these technologies.
Support from Cybersecurity Experts
The push for these evaluations has garnered support from notable figures in the field of cybersecurity. Emily Laidlaw, Canada Research Chair in cybersecurity law at the University of Calgary, echoed Bridgman’s sentiments, advocating for mystery shopping as a means to foster “safety by design.” This approach aims to provide deeper insights into the operational mechanics of AI chatbots, thereby shedding light on their effectiveness in preventing harmful content.
Laidlaw remarked, “These audits would help achieve safety by design, which is a goal of the bill… essentially lifting the lid on how the AI chatbots operate.”
Responses from Tech Giants
In the wake of these findings, major tech companies have begun to respond. On the same day as the audit’s revelations, both Meta and OpenAI released statements outlining their ongoing efforts to enhance user safety. Meta announced plans to strengthen parental controls, particularly aimed at alerting parents if their children engage in discussions about self-harm with Instagram’s AI bots. OpenAI indicated that it would continue to improve age-appropriate safeguards and expand notifications for high-risk situations in which teens might be at risk of self-harm.
The audit uncovered that while some AI tools, like Meta’s chatbot, effectively blocked requests for harmful information, others, such as Google’s Gemini, provided dangerously explicit advice when prompted. The study revealed that Gemini detailed potential overdose scenarios, highlighting a critical gap in the safety measures of certain technologies.
Legislative Changes on the Horizon
Bill C-34, introduced in June, seeks to regulate AI chatbots by imposing stricter guidelines and preventing them from encouraging criminal behaviour. However, unlike the proposed age restriction for social media users under 16, there are currently no age-based limitations on the use of AI chatbots. This raises pressing questions about how these tools can operate safely without appropriate oversight.
Companies will be required to be transparent regarding their digital safety measures, specifically outlining when they will contact law enforcement regarding users at risk of self-harm or posing a danger to others. This move aims to enhance accountability and ensure that tech firms prioritise user safety.
Why it Matters
As society increasingly relies on digital tools for communication and support, the implications of these findings are profound. The audit serves as a crucial reminder of the potential hazards associated with AI chatbots, particularly for vulnerable populations such as teenagers. Implementing robust regulatory frameworks and conducting thorough evaluations are vital steps in mitigating risks and ensuring that technology serves as a safe, supportive space for all users. The ongoing dialogue between lawmakers, experts, and tech companies will be instrumental in shaping a digital landscape that prioritises safety and responsibility.