Skip to content
All AI news
EveryoneProfessionalsParents

AI researchers trick chatbots into sharing how to make cocaine as long as they believe a user is wearing a green shirt — 'CoT Forgery' exploit spurs LLMs to divulge forbidden info by faking trusted chains of thought - Tom's Hardware

Tom's Hardware Jul 1, 2026 at 3:00 AM PDT
0

Researchers discovered that hackers can trick AI chatbots into providing dangerous information by exploiting reasoning vulnerabilities. This reveals a serious security flaw that affects all users relying on these systems.

Read the full article
Share
Discussion

0 comments