The UK Government’s adviser on terrorism legislation, Jonathan Hall KC, has proposed new laws to curb the use of AI chatbots by extremists to radicalise would-be terrorists. In his annual report, Hall suggests targeting computer programmes or chatbots designed to “stir up hatred” against specific groups—namely Jews, Muslims, Black people, and gay individuals. While countering AI-driven radicalisation is a valid concern, Hall’s proposals are dangerously narrow, failing to address the UK’s most significant terrorist threat: Islamist extremism, which poses a far greater danger to the general public, including those outside the communities Hall highlights.
Hall’s report outlines seven ways AI could be exploited by terrorists, from creating propaganda to facilitating attacks. He argues for new offences to tackle chatbots or generative AI that promote hatred based on race, religion, or sexuality. “This would capture Gen AI that was designed to create propaganda intended to stir up hatred against Jews or Muslims or black people or gay people – which probably covers the bases for most terrorist propaganda,” Hall states.
This claim is alarmingly misguided. By focusing solely on these specific groups, Hall’s proposal overlooks the overwhelming threat of Islamist terrorism, which targets the general public indiscriminately, including those who do not fall under the mentioned communities. Islamist extremism has consistently been the UK’s deadliest terrorist threat, dwarfing other forms of terrorism in scale and impact. From the 7/7 London bombings to the Manchester Arena attack, Islamist-inspired attacks have killed and injured hundreds, with victims spanning all demographics but predominantly affecting the broader population in public spaces.
Home Office data underscores this reality. In 2024, 63% of terror-related arrests in the UK were linked to Islamist ideologies, and between 2010 and 2023, the majority of disrupted terror plots were Islamist-inspired. These attacks, often targeting crowded public areas like transport hubs or concert venues, pose a far greater risk to the general public than the niche scenarios Hall’s laws address. Yet his framework fixates on AI chatbots stirring hatred against specific minorities, ignoring the broader, more lethal threat of Islamist radicalisation that endangers everyone, including Jews, Muslims, Black people, and gay individuals.
Consider the case of Jaswant Singh Chail, cited by Hall, who was jailed for nine years after attempting to assassinate Queen Elizabeth II in 2021, partly inspired by a chatbot. This incident, while serious, pales in comparison to the scale of Islamist terror plots, such as the 2017 London Bridge attack, where eight people were killed by attackers shouting “This is for Allah.” These attacks target the public at large, yet Hall’s proposals fail to account for the indiscriminate nature of Islamist terrorism, which poses a greater risk to all communities, including those he seeks to protect.
Hall’s narrow focus suggests either a profound lack of understanding or deliberate avoidance of the Islamist threat. His assertion that targeting hatred against specific groups “covers the bases for most terrorist propaganda” is not only incorrect but dangerously reductive. Islamist extremism thrives on a broader anti-Western narrative that vilifies democratic societies as a whole, often propagated through human networks, encrypted apps, and real-world recruitment. AI could amplify these existing channels, but Hall’s obsession with chatbots targeting specific minorities misses the bigger picture.
Furthermore, Hall’s reluctance to broaden terrorism legislation—claiming it should avoid “opening the door of terrorism liability too far”—is a weak excuse. By limiting protections to specific groups, the proposal dismisses the vulnerability of the general public, who face the brunt of Islamist terrorism’s indiscriminate violence. This selective approach reflects a broader trend in policymaking, where progressive priorities overshadow the need to confront the UK’s most pressing security threat.
Hall acknowledges AI’s potential for “social degradation” through conspiracy theories or grievance narratives that foster violence, yet he dismisses the role of terrorism legislation here because the link to terrorism is “too indirect.” This is an unacceptable cop-out. The radicalisation pipelines fuelling Islamist terrorism—from online sermons to encrypted messaging groups—are well-established and far more dangerous than hypothetical AI chatbots targeting specific groups. AI could exacerbate these threats, but Hall’s proposals fail to address the root ideologies driving Islamist extremism.
The UK deserves a more robust response. Policymakers like Hall must confront the reality of Islamist terrorism, which poses a far greater threat to all communities—Jews, Muslims, Black people, gay individuals, and the general public—than the scenarios his laws target. Ignoring this in favour of a politically correct focus on AI-driven hatred against specific groups is shortsighted and reckless.
If the Government is serious about protecting its citizens, it must broaden its approach to tackle all forms of extremist propaganda, particularly the Islamist ideologies that have claimed the most lives. AI is merely a tool; the real danger lies in the ideologies it could amplify. Hall’s proposals, while well-meaning, are a step in the wrong direction, leaving the nation vulnerable to its greatest terrorist threat.
