Ensuring National Security in the Age of AI-Driven Extremism
About a month ago, Daveed Gartenstein-Ross, Senior Advisor on Asymmetric Warfare at the Foundation for Defense of Democracies, testified before the House Homeland Security Subcommittee on Counterterrorism and Intelligence. During the hearing, he urgently encouraged legislators to address the fastest growing cyber threat the United States is facing: cyber terrorists using generative artificial intelligence for recruitment and mass radicalization.
Generative AI is an actively advancing technology that can generate unlimited personalized content, made up of text, images, or audio. Terrorists and violent extremist groups are now using this tool to influence, brainwash, and radicalize digital audiences with disturbing precision. During the House Homeland Security Committee hearing on online radicalization and terror recruiting led by Chairman August Pfluger, Gartenstein-Ross outlined some notable developments: extremists can now, using AI, produce hyper-personalized content tailored to specific ideological inclinations; they can create deepfake recruitment videos that appear credible; and they are beginning to employ AI chatbots to personally engage potential recruits. Fused with the limitless nature of AI, these escalating capabilities extend the reach, rate, and psychological damage caused by digital adversaries.
To address this threat, Gartenstein-Ross emphasized the proposed “Generative AI Terrorism Risk Assessment Act,” which would require the Department of Homeland Security (DHS) to conduct yearly assessments of terrorism risks related to generative AI tools. In tandem, Chairman Pfluger introduced the “Countering Online Radicalization and Terrorism Act,” a bill that would mandate DHS to keep a close eye on threats posed by terrorist use of foreign cloud-based messaging apps like Telegram, an encrypted messaging platform commonly exploited to circulate extremist content and expand recruitment.
The exploitation of AI tools by extremist factions to mass-produce propaganda, disseminate disinformation, and fabricate content to incite violence or fear is no longer just an imminent risk—it is a reality. AI tools make it easier than ever for domestic and foreign adversaries to exploit vulnerable audiences, threaten national security, and ignite fiery division spanning across borders and platforms.
Digital terror activity has reached unprecedented levels, with groups now publishing how-to guides on using AI for propaganda and digital leverage. In 2023, a pro–al-Qaeda outlet produced multiple propaganda posters with imagery generated using AI. Pro–Islamic State actors have used AI to translate Arabic-language ISIS propaganda into Indonesian and English, diversifying their online audience reach and influence. Further disturbing, an al-Qaeda–affiliated group announced it would begin hosting online AI workshops and released a detailed guide on using AI chatbots for radicalization.
For long-time observers of ISIS, the use of AI by terror groups is no surprise. When ISIS rose to prominence around 2014, it was already utilizing high-production-value videos to threaten enemies and recruit new members. With AI, those digital capabilities are boundless.
Today, the primary targets of these efforts are young adults. Social media sites and encrypted messaging platforms have made it easier than ever for extremist groups to reach impressionable audiences with tailored, ideologically loaded content. Generation Z’s ease with technology and constant online presence makes them particularly vulnerable to falling victim to online radicalization tactics.
In response to the growing digital terror network, Big Tech outlets such as Microsoft, Google, and Facebook have collaborated to create an active interconnected hashing database of known violent extremist content, designed to immediately and automatically remove extremist content across platforms.
Although AI has complicated and amplified the extremist cyber security risks, Gartenstein-Ross and Chairman Pfluger have their fingers on the pulse of this issue, sounding the alarm on the threats posed by the uptick of AI-driven radicalization. Just as quickly as cyber terrorists can reach mass audiences and AI can generate inciteful propaganda, disaster can unfold with irreversible consequences. Extensive surveillance and online monitoring is mandatory to prevent AI technologies from being exploited to further terror initiatives. It is crucial that cybersecurity and counterterrorism measures remain a bipartisan issue, and that urgent action is taken to implement consistent oversight of extremist activity online.
