• Artificial intelligence (AI) will continue to have wide ranging, transformational, and often beneficial, impacts on society. However, like all technology, AI also carries the potential for misuse by nefarious actors including Terrorists and Violent Extremists (TVEs).
•  Though this technology continues to develop at a rapid pace, understanding how existing AI models work can help assess how and where they are likely to be beneficial to TVEs.
• AI presents various tools TVEs might employ to streamline, support, and speed up propaganda production. However, the ability of this technology to mass produce impactful propaganda without human oversight is likely subject to several limitations.
• AI is likely to have significant impacts for the dissemination of TVE propaganda, providing new means to amplify content and circumvent the automated detection and removal systems used by social media platforms.
• Whilst there are potential radicalisation risks associated with AI, including those stemming from AI-powered chatbots, there is currently limited evidence to assess the nature and extent of this risk.
• AI is likely to offer significant benefits to support TVE’s operations and activities, including the planning, facilitation, and execution of violent attacks.
• TVE exploitation of AI is currently in an experimental phase. There is currently little evidence to suggest a widespread and transformative adoption of these tools in the immediate future. Rather, TVE uptake is likely to be incremental and will be dictated by multiple factors including these actors’ own perceptions of this technology as well as changing societal and counterterrorism uses of AI.

Significant attention has been devoted to predicting and understanding how artificial intelligence (AI) will reshape various aspects of human life. Though AI has already provided large and transformational societal benefits (including counterterrorism and threat detection {1}), like all technology it also carries the potential for misuse by Terrorists and Violent Extremists (TVEs). {2} Though potential TVE exploitation of AI will overlap with those of other nefarious actors (for example, AI-facilitated scams, fraud, or other forms of cybercrime for fundraising), other uses are likely to be unique and inherently linked to the goal of advancing a political, religious, racial, or ideological cause.

 Whilst much has been written about the significant opportunities AI could provide to TVEs, it is necessary to consider the factors likely to shape these actors’ capacity or willingness to embrace these opportunities. Teasing apart instances where AI might have a transformative impact, providing TVEs with new capabilities, and where it may merely speed up or make their operations more efficient may also be useful for assessing potential threats. Though AI continues to develop at a rapid pace, systems with human-like intelligence, that extend beyond the ability to perform very specific tasks, (referred to as ‘Artificial general intelligence’) remains confined to science fiction. Rather, contemporary AI models are essentially statistical engines capable of identifying patterns in training data before using this knowledge to generate new, unique outputs. As such, though these models can learn to solve problems at speed and scale impossible for humans, they do not understand the wider contexts of requested tasks, nor do the meanings of the outputs they generate