Gmail AI Assistant: Convenient But Potentially Compromised
Exploring security vulnerabilities in Gmail AI features including indirect prompt injection and link trap risks, and the stance of Google on these issues
January 4, 2025
Read more
Why "Following the Science" in Cybersecurity Is Misguided
An analysis of why strictly following scientific approaches in cybersecurity can be problematic and may not account for real-world complexity
December 19, 2024
Read more