Research conducted by the Cambridge Programme on AI Science & Policy reveals that factions of Boko Haram have begun integrating publicly available artificial intelligence systems into their operations, marking a significant shift in their tactical approach.
AI integration spans beyond propaganda
While AI has traditionally been associated with content creation for propaganda purposes—such as generating texts, images, or videos—this study highlights far more sophisticated applications. Former mid-level commanders and technical specialists from the group’s splinter factions have disclosed how AI tools are now being used to plan operations, resolve logistical challenges, and enhance coordination between units.
These factions, which include the Islamic State’s West Africa Province (ISWAP) and Jama’atu Ahlis Sunna Lidda’awati wal-Jihad (JAS), have reportedly employed AI platforms such as ChatGPT, Claude, Gemini, Grok, Meta AI, and DeepSeek since 2023. Their usage extends to analyzing failed attacks, optimizing future operations, and even troubleshooting captured military equipment.
Structured AI cells and Daesh influence
Some units have gone further by establishing dedicated AI cells composed of members with expertise in engineering, weaponry, explosives, and strategic planning. These teams act as intermediaries between field operatives and AI systems, querying multiple platforms to extract and refine actionable intelligence.
Restricted access to computers, accounts, and subscriptions ensures operational security, minimizing risks of surveillance or unauthorized use. The adoption of these tools has reportedly been encouraged by instructors linked to Daesh, who provided in-person training, remote assistance, and even supplied hardware such as laptops, encryption software, and virtual private network (VPN) access.
Technological limitations and security concerns
The study, based on interviews with 27 former members between 2025 and 2026, acknowledges methodological constraints, as conclusions rely solely on their testimonies. While AI has accelerated access to technical knowledge—reducing research time from days to minutes—its outputs are not always accurate or applicable in real-world scenarios.
Moreover, the report warns that while AI democratizes access to advanced tools, it also complicates efforts to detect malicious use. The proliferation of platforms, accounts, and circumvention techniques poses challenges for developers, security services, and public authorities alike.
Authors urge stronger collaboration among stakeholders to evaluate safeguards against structured adversaries rather than isolated users. They emphasize that the question is no longer whether armed groups will adopt AI, but how effectively security mechanisms can curb its misuse as a tool of violence.
