Labour to bring in new law to stop violent extremist groups using AI for radicalisation and terror threats

In a major new speech at RUSI today Monday 17th July, the Shadow Home Secretary will set out Labour’s approach to national security, including action to stop violent extremist groups using AI to radicalise vulnerable people online. 

Yvette Cooper will argue that our intelligence and security agencies are among the best in the world and are leading the way but that the Government is being outpaced by new and emerging threats the country faces. She will say that she is determined to restore a sense of purpose and leadership to the Home Office on national security. 

The Shadow Home Secretary will point to a series of new threats to national security, arguing that the Government is not sufficiently agile in responding to a fast-changing security landscape. In particular, she will say that action is needed to tackle the terror risks associated with the rising use of AI technology, including the use of chatbots by extremist groups to radicalise users and disseminate terrorist propaganda.  

Whilst deliberate instigation of terror acts is a criminal offence under terrorism legislation, establishing criminal liability is less clear an individual has been radicalised through AI - even when AI systems have been trained for that purpose. Cooper will announce that Labour will close this loophole, criminalising the deliberate training of chatbots to spread terrorism and violence, and working with the intelligence community to stop malicious actors. The announcement comes following the sentencing of Jaswant Singh Chail, who attempted to carry out a terror plot to kill the Queen at Windsor Castle with a loaded crossbow, having been encouraged by his ‘AI girlfriend’ to proceed with the attack. 

She will say: 

“Artificial Intelligence creates new opportunities for Britain, including for law enforcement, but it also presents significant new threats and risks. A series of recent cases have revealed the potential for online chatbots to be used to radicalise people with pro-terror content. 

“Our law enforcement and legislation must not be outpaced by terrorists and extremists using new technologies to prey on vulnerable people. 

“That’s why Labour will criminalise those who purposely train chatbots to spout terrorist material, with stronger action to monitor and stop radicalising chatbots that are inciting violence or amplifying extremist views.”

In her speech on Monday, the Shadow Home Secretary will say that the Government’s CONTEST update on counter-terrorism due this week must include action to tackle hateful extremism, terrorist finance, the rise in online radicalisation, and the deliberate misuse of AI by organised extremist groups.

But she will also warn that a new cross-government strategy for state threats is needed to run alongside the counter-terror CONTEST plan. She will say that a new Labour Government would address the increasingly blurred space between terrorism and conventional state threats, particularly around the use of violence, repression and interception on UK soil to illicitly advance the interests of foreign regimes or crack down on dissidents.

Cooper will say that Labour will look to rebuild the consensus about the importance of defending our national security, arguing that this was undermined by behaviour such as Boris Johnson’s decision as Foreign Secretary to meet with a former KGB officer, Alexander Lebedev, at the height of concern around the Skripal poisoning.  

Ends

NOTES

• Labour will legislate to criminalise the deliberate training of chatbots to spread terrorism and support the intelligence community to stop malicious actors.
• In June the government's Independent Reviewer of Terrorism Legislation Jonathan Hall KC warned of the lack of legal clarity around AI-directed terrorism: "Again, it is unclear how legal culpability would be established in a scenario where an individual was radicalised (in part) by an AI system.” https://cetas.turing.ac.uk/publications/terrorism-and-autonomous-weapon-systems-future-threat-or-science-fiction
• In consultation with experts, the Shadow Home Secretary will announce that a Labour Government will close this loophole, revising the scope of offences covered by both the Terrorism Act 2000 (TACT), and the Terrorism Act 2006.
• The announcement comes following the sentencing of Jaswant Singh Chail, who attempted to carry out a terror plot to kill the Queen at Windsor Castle with a loaded crossbow. It emerged that the convicted terrorist had been encouraged by his ‘AI girlfriend’, developed through the real-life imitation chatbot app Replika, to proceed with the attack. 
• Experts and academics have been warning about the potential malicious use of artificial intelligence (AI) by rogue states, criminals and terrorists, for many years. 
• Just last week the US Federal Trade Commission officially launched an investigation into AI chatbots, after announcing a wide-ranging probe into ChatGPTmaker and OpenAI.