Boko Haram, AI and "the white guys"
MuckyNews reading
Boko Haram has been using AI chatbots to build bombs, plan attacks, and spread propaganda. That's the core finding of a new Cambridge University study based on interviews with 27 former fighters. The group apparently used both US models, like ChatGPT and Claude, and the Chinese provider DeepSeek. The real oddity here is how they learned to use the tech: external consultants — dubbed "the white guys" by the fighters — reportedly ran workshops showing members how to operate the tools using laptops set up with VPNs and encryption software. These trainers were likely part of the broader ISIS network.
So, what does this tell us? AI is no longer just a toy for tech nerds; it’s operational infrastructure for militias. The dividing line isn't between high-tech states and low-tech insurgents anymore, but between those who have access to these tools and those who don't. The pressure point is obvious: if jihadist networks are already running AI workshops for local militants, relying on voluntary corporate safeguards won't cut it. The study calls for the US and China to urgently address the misuse of AI by non-state actors in their upcoming safety talks. A reasonable demand — though aimed at two players who are usually more focused on out-competing each other in the tech race.
Short version
- What happened? A Cambridge study shows Boko Haram used US and Chinese AI
chatbots for bomb-making, attack planning, and propaganda.
- Who is involved? Boko Haram, external ISIS-linked AI trainers ("the
white guys"), OpenAI, Anthropic, DeepSeek.
- Why does it matter? Terrorist groups are treating off-the-shelf AI as
operational tools and getting schooled by jihadist networks on how to use them.
- What is still open? It remains unclear whether the AI providers noticed
or prevented this usage, or what the direct impact on specific attacks was.
- What is the real pressure point? AI has become a standard resource
for non-state violent actors, putting the burden on the US and China jointly to address it.
Perspective delta
- **hongkong_china_asia**: The reporting (SCMP) heavily emphasizes shared
responsibility between the US and China. Both superpowers produce the models being used, and the study argues both must address this in their AI safety talks. By placing Chinese and American tools squarely side-by-side as equal parts of the problem, the framing distributes the blame symmetrically. What’s left out is any discussion of whether the Chinese government is already regulating DeepSeek more strictly domestically than the US does with its tech firms.
Source base / limits
The source base is extremely narrow. The material relies on a single article from the South China Morning Post, which itself is based on one research paper. The primary data comes from interviews with just 27 former Boko Haram fighters. There are no statements from the AI companies involved and no independent verification from other media outlets. The perspective is strictly academic and filtered through an Asian lens, and the text base is thin. It also remains unclear how representative the statements of those 27 ex-fighters really are for the group as a whole.
Sources / basis
- MuckyNews container ID: MN-20260714-078
- Source outlets: South China Morning Post (SCMP)
- Perspective groups: hongkong_china_asia
- Article count: 1
- Source count: 1
- Perspective count: 1
- Approximate raw text size: 2186 characters / 546 tokens