UN Report Warns Terrorists Are Using AI to Radicalise, Recruit and Incite Violence

Terrorist groups are increasingly turning to artificial intelligence to radicalise, recruit and incite violence, according to a new United Nations report that warns the technology is being used to generate propaganda at unprecedented speed and scale. The report, titled “CTED Trends Report Beyond messaging: terrorist propaganda as operational infrastructure,” was published by the UN Counter-Terrorism Committee Executive Directorate (CTED) and found that extremists are experimenting with generative AI—systems capable of creating text, images, audio, video and code in response to a user’s prompt. Widely available tools such as ChatGPT, Google’s Gemini and Claude are among the platforms that can be exploited for terrorist purposes. The report says that terrorist and violent extremist actors are using these tools to produce and distribute propaganda more efficiently than ever before, influencing interactive recruitment and behaviour through social media channels, spreading disinformation and ideologies conducive to terrorism, and inciting violence. It also warns that terrorist groups are issuing guidance to supporters on how to evade detection while using generative AI tools, making it harder for authorities to monitor their activities. The report highlights that AI applications can unintentionally accelerate the radicalisation of vulnerable individuals as users increasingly rely on them for companionship, guidance and advice. Recent research, it notes, suggests that AI may function as a radicalisation environment rather than merely a tool, potentially reinforcing grievances, identities and existing beliefs through sustained personalised interactions that remain largely invisible to traditional counter-terrorism monitoring approaches. The findings come as governments around the world struggle to keep pace with rapidly evolving AI capabilities and their security implications, and as the line between human-generated and machine-generated extremist content becomes increasingly blurred.

Generative AI’s ability to translate content into multiple languages simultaneously is particularly concerning, the report says. Terrorist groups can now reach audiences across the globe in their own languages, enabling rapid outreach and disinformation campaigns that feed radicalising content to unsuspecting users. At the same time, personalised AI interactions are being exploited to reinforce existing beliefs and vulnerabilities, heightening susceptibility to terrorist influence and recruitment. The report observes that AI may further amplify the role of gender in terrorist propaganda, with experts cautioning that generative AI enables the rapid production and localisation of gendered propaganda tailored to men and women, girls and boys. This allows terrorist and violent extremist actors to adapt narratives to specific audiences at greater scale and to disseminate them at an accelerated pace. The report also points to groups motivated by xenophobia, racism and other forms of intolerance, or in the name of religion or belief (XRIRB), as having similarly operationalised AI for propaganda production. These groups are using AI not only to spread hate but also to build what the report calls operational infrastructure, using propaganda as a means to coordinate activities, recruit new members and fundraise. The report’s authors stress that AI is not merely a communications tool but an operational one, embedded in the broader infrastructure of terrorist activity. By enabling the rapid creation of tailored content across multiple languages and platforms, AI allows extremist groups to maintain a constant and adaptive presence online, reaching vulnerable individuals who might otherwise never encounter their propaganda. The report warns that this represents a significant escalation in the sophistication and intensity of terrorist propaganda, with potentially far-reaching consequences for public safety and social cohesion.

At the operational level, the report documents specific uses of AI across groups and regions. It references findings suggesting that Da’esh, also known as ISIL, has sought to recruit cyber experts to bolster its capacities and tailor propaganda to local contexts. Da’esh supporters have launched an AI-generated programme to disseminate propaganda videos, the report says, demonstrating how the technology is being used to automate and expand outreach efforts. The report further underscores that AI-enabled propaganda is increasingly intersecting with terrorist financing. It notes that ISIL-K, the Islamic State’s Khorasan Province affiliate, has integrated cryptocurrency fundraising directly into its propaganda. More broadly, large language models have the potential to lower barriers to fundraising by enabling the rapid generation of tailored, multilingual narratives and coordinated outreach campaigns, allowing terrorist actors to expand donor engagement and enhance perceived legitimacy at relatively low cost. This combination of AI-generated propaganda and cryptocurrency creates new challenges for financial investigators and counter-terrorism officials, as traditional methods of tracking terrorist financing may be less effective against these emerging techniques. The report’s findings suggest that terrorist groups are adapting quickly to technological changes, using AI to enhance both the sophistication and intensity of attacks and to incite violence. The use of AI also allows these groups to operate with greater agility, producing content that is specifically designed to resonate with different audiences, whether based on language, culture, age or gender. This operational flexibility makes it more difficult for counter-terrorism authorities to anticipate and disrupt terrorist activities, as the volume and variety of AI-generated content can overwhelm monitoring systems and human analysts.

The report underscores significant challenges for Member States in detecting and countering AI-enabled terrorist content. It says that states have frequently reported difficulties identifying, monitoring and countering terrorist content produced outside their jurisdictions, often in foreign languages and disseminated through unfamiliar or rapidly changing platforms. The growing use of AI by terrorist groups further exacerbates this gap. These dynamics can leave vulnerable communities exposed to large volumes of propaganda while limiting the ability of states to detect, disrupt or offer credible alternatives to such messaging, the report adds. The cross-border nature of AI-generated content complicates law enforcement efforts, as does the sheer volume of material that can be produced with minimal human effort. Unlike traditional propaganda, which requires time, resources and human creativity, generative AI can produce endless variations of text, images and videos in a matter of seconds, making it nearly impossible for platforms and authorities to keep up. Moreover, AI systems can be used to create content that is tailored to the specific interests, fears and grievances of individual users, increasing the likelihood that it will be accepted and shared. The report warns that this can create echo chambers in which vulnerable individuals are gradually drawn into extremist ideologies without ever coming into contact with counter-narratives or alternative viewpoints. It calls for greater international cooperation and information sharing to address these challenges, while also recognising the need to protect human rights and fundamental freedoms in any counter-terrorism measures. The report’s authors argue that the threat posed by AI-enabled terrorist propaganda cannot be addressed by any single country alone, and that a coordinated global response is essential.

In response to these threats, a growing number of Member States have adopted or are considering legislative and regulatory measures to restrict young people’s access to certain online services and social media platforms, aiming to reduce exposure to harmful content and online risks. However, the report notes that such efforts have also attracted litigation and criticism from human rights agencies in instances where users’ right to privacy is infringed upon. This highlights the delicate balance between security and civil liberties in the digital age. The report also identifies emerging good practices that could help counter AI-enabled terrorist propaganda. These include the expansion of multidisciplinary intervention programmes for individuals influenced by mixed, hybrid, or ideologically fluid forms of violent extremism, and the development of off-ramping mechanisms that connect at-risk online users with offline support services. Such approaches recognise that countering radicalisation requires more than just removing content; it requires engaging with vulnerable individuals and providing alternatives to extremist narratives. The report suggests that governments, technology companies and civil society must work together to develop effective responses that respect human rights while addressing the evolving threat. It also emphasises the importance of media literacy and digital resilience, helping young people and other vulnerable groups to critically evaluate online content and resist manipulation. The report’s recommendations reflect a growing consensus that counter-terrorism strategies must evolve in tandem with technological developments, and that a purely reactive approach is no longer sufficient. By combining legal measures, technological tools and social interventions, Member States can better protect their citizens from the dangers of AI-enabled radicalisation while upholding democratic values and the rule of law.

The CTED report serves as a stark reminder that artificial intelligence is a double-edged sword. While generative AI offers immense benefits for society, it also provides terrorist groups with powerful new tools for propaganda, radicalisation and recruitment. The report’s findings indicate that AI is already being used by a range of actors, from Da’esh to xenophobic and racist groups, and that the threat is likely to grow as the technology becomes more accessible and sophisticated. The UN report calls on Member States to enhance their capacities to detect and counter AI-enabled terrorist content, while also addressing the underlying conditions that make individuals vulnerable to radicalisation. It emphasises the need for international cooperation, public-private partnerships and continued research into the ways AI is being exploited by terrorists. At the same time, it cautions against overbroad responses that could infringe on human rights and privacy, noting that effective counter-terrorism must be grounded in the rule of law. As AI continues to evolve, so too must the strategies of those seeking to prevent terrorism and violent extremism. The report makes clear that the challenge is not merely technological but also social, political and legal, requiring a comprehensive and coordinated global response. The coming years will likely see an intensifying contest between those who use AI to spread hate and violence and those who use it to protect communities and uphold democratic values. The international community must act now, the report argues, to ensure that the same technology that powers innovation and progress is not allowed to become a weapon of mass radicalisation. Only through vigilance, cooperation and a steadfast commitment to human rights can the world hope to stay ahead of this emerging and rapidly evolving threat.