UK Mandates 48-Hour Deepfake Takedowns
A new UK law requires tech firms to remove non-consensual intimate images, including AI-generated deepfakes, within 48 hours of notification. Companies that fail to comply with the rapid takedown timeline will face substantial fines, establishing a new legal standard for platform accountability.
- This mandate is an extension of the UK's Online Safety Act, which designates the creation and sharing of non-consensual intimate images as a "priority offence," granting the regulator Ofcom the power to enforce the rapid takedown. - Penalties for non-compliance are severe, with potential fines reaching up to 10% of a company's global annual revenue or, in extreme cases, having their services blocked entirely within the UK. - The legislation closes a significant loophole by criminalizing not only the sharing of deepfakes but also the act of creating or commissioning them, even if the images are never distributed. - The law was fast-tracked by the government partly in response to public outrage over the use of AI chatbots on platforms like X (formerly Twitter) to generate abusive deepfake images, which triggered a formal investigation by Ofcom into the platform. - As a next step, the UK government is amending its Crime and Policing Bill to explicitly ban the development and supply of "nudification" apps and tools designed to generate non-consensual intimate images. - Internationally, the UK's approach of fining up to 10% of global turnover is stricter than the EU's Digital Services Act, which allows for fines of up to 6%. However, other jurisdictions like Türkiye are proposing even faster timelines, with a 6-hour takedown requirement for certain harmful content. - Victims can report images directly to tech firms or to Ofcom, which can then trigger an alert across multiple platforms to prevent the same image from being repeatedly reposted, a system that mirrors the "hash matching" technology used to combat the spread of child sexual abuse material.