Terrorist Groups Have Created Dedicated "AI Units"

A research program at the University of Cambridge studying the use of AI in political contexts has found that militants from two factions of Boko Haram and ISWAP terrorist organizations are systematically using ChatGPT, Claude, Gemini, Grok, Meta AI, and DeepSeek.

According to former members of the groups, these AI systems are used not only for propaganda, but also for operational planning, weapon repair, bomb-making, security, and analyzing previous attacks.

The militants reportedly bypass chatbot safety restrictions by rephrasing their prompts or switching between different AI models. The report's author, Dr. Antonia Julich, calls on AI companies, intelligence agencies, and researchers to share information about such cases more actively.

She also argues that AI safety testing should focus not only on large-scale attack scenarios but also on everyday tasks that can gradually make terrorist organizations more effective.