Some users recommend the full paper on AI-enabled terrorism because it offers in-depth analysis of reports about Boko Haram using AI chatbots and models for bomb designs and tactics.
Based on 1 visible X reactions from 4 accounts; directional sample.
Ask a question below.
Published answers will appear here.
I can't recommend the full paper enough: https://casp.ac/reports/ai-enabled-terrorism
For years, the AI risk community has highlighted the potential for misuse of more capable AI by cyberattackers and non-state actors. These concerns received pushback as being unproven and sensationalist Mythos demonstrated the cyber concerns were well-founded. Now Antonia Juelich has demonstrated in no uncertain terms that concerns around terrorist use of AI is similarly well-founded. Boko Haram have used ChatGPT, Gemini, Claude, DeepSeek etc routinely across a wide range of functions. This is hugely important work by Juelich, plugging a critical evidence gap with admirable rigor. No small amount of bravery either (getting funders to cover her ransom insurance was not a small challenge). Two further observations: 1) This might encourage us to look again at other concerns currently considered 'unproven' or 'sensationalist', including AI-bio and loss of control. Catastrophic harms aren't guaranteed here, but the world can't afford to be caught wrong-footed. 2) As evidence mounts, these challenges present opportunities. Non-state actor misuse of AI is not in any leading nations' interests. Even if competition dominates elsewhere, there is ample room, and even motivation, for cooperation between the US, Europe and China on these threats, and active interventions that we can cooperate on. Multiple workshops I've been at with Chinese, European and US research leaders have convinced me of this.
Antonia told me about this paper a couple weeks ago, and it blew my mind. Boko Haram terrorists are using frontier AI models — including ChatGPT and Claude — to plan attacks, troubleshoot weapons, and design explosive devices. Islamic State operatives gave them in-person training on how to use AI tools, and the AI tools appear to providing real-world uplift. This is all based on interviews with 27 former Boko Haram members. Terrorist misuse of AI has long been a theoretical risk people have talked about; this new paper shows it's now reality.
@S_OhEigeartaigh the leading "use" of AI mentioned here, ah, seems pretty net-negative for the terrorists they had 18 deaths... after getting the AI to help them jump motorcycles over trenches? AI helping them execute literally movie-inspired tactics? unless they were just lying to researchers https://x.com/1a3orn/status/2075604618005094512/photo/1
@S_OhEigeartaigh some of these things seem genuinely dual-use, but in a way I'm quite comfortable with I want AI to be able to help people encrypt things, for the same reason I want gpop to be able to access industrial grade encryption https://x.com/1a3orn/status/2075604998176854483/photo/1
Former commanders fed combat data into ChatGPT, Gemini, and Claude
@S_OhEigeartaigh some of these things seem genuinely dual-use, but in a way I'm quite comfortable with I want AI to be able to help people encrypt things, for the same reason I want gpop to be able to access industrial grade encryption https://x.com/1a3orn/status/2075604998176854483/photo/1
The examples in the paper are wild. "When ISWAP fighters were handed out new guns and they did not know how to correctly use them, they approached their qaid, who passed on the message to a specialist, who in turn replied, 'just ask Grok,' which they then did.
Jihadist terrorist groups might be better at adopting AI than most US companies 🤔 https://twitter.com/AntoniaJuelich/status/2075590815083028989
The report is available here: https://casp.ac
@1a3orn Life is cheap to these people.
Some users recommend the full paper on AI-enabled terrorism because it offers in-depth analysis of reports about Boko Haram using AI chatbots and models for bomb designs and tactics.
Based on 1 visible X reactions from 4 accounts; directional sample.
Ask a question below.
Published answers will appear here.
The examples in the paper are wild. "When ISWAP fighters were handed out new guns and they did not know how to correctly use them, they approached their qaid, who passed on the message to a specialist, who in turn replied, 'just ask Grok,' which they then did.