ISD’s Head of Policy and Research, Milo Comerford, spoke to iNews about the potential abuse of generative AI by extremists. While noting that extremist innovation using AI tools is still in its infancy, Milo discussed the potential application of these tools to overwhelm platforms through coordinated online campaigns. He highlighted how generative AI could serve as a tool to increase engagement with propaganda while lowering the barriers and costs of translation and content production.
Milo explained how extremists could exploit AI to launch coordinated campaigns by producing similar or identical messages to flood social media platforms which consequently can “increase the scale and engagement of extremist propaganda.”
“ISIS and other groups are recruiting all around the world (…). [AI tools] are often better than Google Translate, and generative AI allows you to be more idiomatic and adopt a more casual style,” he said.
Milo also pointed out that translating messages using AI instead of mainstream tools could help extremists evade detection. However, he also emphasised that the use of AI tools among extremist networks is a “relatively limited phenomenon” and that the use of these tools is “broadly in line with the rest of population”.
The article featured additional commentary from partners at the Community Security Trust who noted a major uptick in AI-generated antisemitic imagery in the wake of the 7 October Hamas attack in Israel.