The party is considering a proposal to outlaw tools used for nonconsensual pornography and has pledged not to tolerate electoral misinformation
Labour is considering proposals to address AI deepfakes, including a ban on ‘nudification’ tools for generating nonconsensual pornography and committing to a cross-party pledge against deepfakes spreading electoral misinformation. The policy paper, from Labour Together, suggests legally banning tools for creating explicit content from real people’s images. It also proposes obligations for AI tool developers and web hosting companies to prevent the production of harmful deepfakes. Co-author Laurel Boxall noted a public attitude towards technology that is close to outright hostility due to past tech scandals like the A-levels results algorithm.
“I believe regulation can be beneficial for AI in this regard: it allows the government to demonstrate to the public that it is supportive and will intervene if technology is being used in harmful ways. It’s important to remember that AI can be beneficial; regulation ensures it doesn’t cause harm as well.”
Her co-author Kirsty Innes noted, “Our proposal goes beyond just nudification, although that is understandably receiving a lot of attention. There are significant risks from deepfakes, such as fraud. Addressing this issue now can help prevent a long-term crisis similar to the rise in online fraud, which is currently the most frequently committed crime in the UK.”
The paper follows reports that Twitter, now known as X, refused to remove a deepfake audio clip of Keir Starmer in October of last year, despite having policies against such misinformation. The AI-generated clip, which falsely depicted the Labour leader criticizing his staff, was quickly removed from other social media platforms. However, according to a Bloomberg report, Twitter declined to take action, citing a policy that allowed them to do so if they were “unable to reliably determine if media have been altered or fabricated.”
Peter Kyle, the shadow secretary for science, innovation, and technology, stated that the party is reviewing the proposals. “The potential for deepfakes to harm individuals, disrupt elections, and escalate fraud has been evident for some time. The recent proliferation of nudification tools is particularly alarming.
“These are advancements with potential for harm. That’s why, during a recent visit, I directly addressed these issues with tech leaders in Seattle and Silicon Valley, as well as policymakers in Washington. I am actively engaging with stakeholders on these matters, and we are carefully considering the proposals.”
The policy paper suggests that all major parties should voluntarily agree not to use deepfake technology or spread misinformation for campaign purposes. It proposes that Labour should join a cross-party pledge “not to create, disseminate, or tolerate misleading information.” Additionally, it urgently requests an exception to the ban on media coverage of a general election on polling day. This exception would allow mainstream media to counter fraudulent misinformation that might be spreading as people go to vote.
