Major security weakness found in leading open AI models

International research team finds all 21 leading open-weight AI models tested could be modified to bypass safety protections, raising concerns about misuse at scale

By Media Relations

Rather than competing on the size of AI models, Canada can lead the global AI revolution by educating engineers and leaders who build trustworthy systems from the ground up. The Waterloo Engineering dean argues that if every graduate gained real-world experience deploying AI safely in organizations, Canada would create one of the world's most trusted and AI-ready workforces.

To read the full article, click here!