Veteran Red Teamer Reflects on Six Years of AI Security Research

Security researcher wunderwuzzi revisited his six-year-old Machine Learning Attack Series, noting his attack-surface assessments still hold. He observed that the most interesting AI vulnerabilities tend to emerge at the intersection of model safety and system security.

2026-10-08 ~ 2026-10-08 · 2 related posts

• Security researcher: the best AI bugs live at the safety-security intersection — wunderwuzzi23 · 2026-10-08
• Six years on, wunderwuzzi's ML attack series on model backdoors and Assume Breach still holds up — wunderwuzzi23 · 2026-10-08