A New York Times report highlights how Boko Haram militants have begun utilizing artificial intelligence to enhance their weapon manufacturing and operational effectiveness. Former fighters reported employing various AI platforms, including ChatGPT, Claude, Gemini, Grok, and DeepSeek, to refine their combat strategies.
Antonia Juelich, a researcher from the University of Cambridge, gathered testimony from ex-commanders who used these tools to overcome military defenses. After a defensive trench stalled a base assault, the group utilized AI to calculate the mechanical modifications needed for their motorcycles to jump over such obstacles. By inputting specific technical data regarding their vehicles and the distances required, they received step-by-step instructions for performance upgrades.
The militants also turned to AI for advice on explosive devices. One former commander described the technology as a human-like resource that provides precise instructions on bomb assembly, noting that it helped them replace dangerous trial-and-error methods with more accurate chemical formulations. These advancements reportedly resulted in more powerful explosions.
While companies like OpenAI, Google, and Anthropic maintain that their models have safeguards against dangerous queries, former insurgents claimed they bypassed these restrictions by masking their inquiries as harmless projects. These tech firms have reiterated that such activities are clear violations of their safety policies.