A report by The New York Times has revealed that Boko Haram insurgents exploited artificial intelligence tools, including ChatGPT, Claude, Gemini, Grok and DeepSeek, to improve their operations and develop more effective weapons. The report, based on interviews with former insurgent commanders conducted by University of Cambridge researcher Antonia Juelich, showed that the group used AI to overcome battlefield challenges and enhance bomb making techniques.
According to the report, Boko Haram fighters turned to AI after an attack on a military base was thwarted by defensive trenches. By providing details about their motorcycles and operational needs, the insurgents reportedly received guidance on modifying the bikes for greater speed and acceleration, enabling them to navigate such obstacles during future attacks.
Former commanders also disclosed that AI chatbots provided detailed responses to questions about explosives, including suggestions on chemical combinations that increased the impact of improvised bombs. They claimed the platforms were frequently used after disguising their requests as legitimate research or engineering projects to bypass built in safety restrictions.
Reacting to the findings, OpenAI said the reported activities violated its usage policies, while Google and Anthropic maintained that their AI models are designed to reject harmful requests. However, the report said former insurgents claimed they were often able to circumvent these safeguards, raising fresh concerns over the misuse of generative artificial intelligence by terrorist groups.