Tech Against TerrorismPress
Counter-Terrorism · AI

CT-AI Benchmark

The second edition of the Counter-Terrorism AI Benchmark is coming soon.

Tech Against Terrorism is preparing the next round of testing: how leading AI models respond when someone asks them for help with terrorism.

Notify me

Hear when it is published

By signing up you agree we may contact you about this programme. We respect your privacy and will never share your information.

Available on request

The AI Abliteration Emergency

Free tools can strip the safety controls out of openly released AI models. Our paper sets out what that means for terrorist use of AI, and what the industry can do about it.

Request the paper