UK cross-party proposal would mandate CSAM detection software on all smartphones sold in Britain

The proposal

A cross-party alliance of MPs and peers is preparing to put plans before the UK parliament that would require technology companies to install detection software on every smartphone and tablet sold in Britain, according to The Observer. The software would automatically blur images and videos of child sexual abuse, including AI-generated material, prevent such content from being created on the device, and make the livestreaming of abuse impossible. Handsets and tablets already in circulation could be brought under the same rules through a mandatory software update.

Andy Burnham told the paper the protections would be "built into the device and switched on by default".

Who is behind it

The campaign is being led in the House of Lords by John Nash, a Conservative peer and former education minister who previously championed Britain's ban on social media for under-16s. In the House of Commons, the drive is being led by Jess Phillips, the Labour former safeguarding minister who resigned over what she saw as the government's slowness to block nudity on children's phones.

Nash told The Observer there was a "ghastly worldwide industry" producing such material and that Britain was one of its biggest consumers, giving the country a responsibility to act. "If it can be done for children's phones on these images, it can be done for child sexual abuse images on all phones," he said, adding: "Frankly, these are trillion-dollar companies."

The route into law

The government has already promised legislation to force Apple and Google to stop children from taking or sharing nude images on their devices. In September, culture secretary Lisa Nandy said the major technology companies had "failed to protect children from some of the worst harms" and that she was determined to bring about "whole ecosystem changes". A government spokesman described the plans as "world-first legislation" within "the strongest online safety laws in the world", alongside new police powers covering criminal use of emerging technologies.

There is a significant gap between the two positions: the government's commitment targets children's devices, while the new proposal would extend detection to every phone in the country. If ministers do not legislate, Nash intends to table an amendment, probably to the forthcoming national security bill, which is expected to include a raft of cybersecurity measures. He says he is confident of passing it in the Lords, where peers previously voted overwhelmingly for the under-16 social media ban.

The technology and the evidence

Supporters argue the required tools already exist. Phillips said software capable of blurring or blocking such images is available today and urged the government to look at what else it could achieve. Simon Bailey, the former chief constable of Norfolk who led the National Police Chiefs' Council's child protection work for eight years, said the technology exists to identify such material "with a degree of confidence".

The Observer also cites figures from the Internet Watch Foundation, which removed more than 312,000 online child sexual abuse images last year, a 7% increase on 2024, and found 3,440 AI-generated videos of child sexual abuse, compared with only 13 in 2024. A paper by technology, law enforcement, cybersecurity and child protection experts, due to be published this week, proposes a "paradigm shift" toward prevention, using AI tools to identify and disrupt suspected abuse on devices before images are viewed or shared. Its contributors include Mark Bevan, Interpol's head of crimes against children; Dan Sexton of the Internet Watch Foundation; and Sarah Napier of the Australian Institute of Criminology.

Polling by The Observer suggests strong public support. Only 6% of respondents said they were very confident that technology companies are doing enough to detect and remove such material; after a short description of the blocking technology, 80% supported installing it on their own smartphone or tablet, 5% opposed it, and close to 90% of both Labour and Conservative voters were in favour.

Why it matters

The proposal shifts content moderation from platforms and servers into the operating system itself. Software that must inspect images and video before they are created, viewed or streamed runs on the device and continuously classifies what users see, including material sent through end-to-end encrypted services, which is only ever readable on the endpoints and never on a company server. Detection therefore cannot happen in transit; it has to happen inside the handset, which is exactly where cryptographers and security engineers have objected to client-side scanning in the past, arguing that infrastructure built to inspect one category of content can be repurposed and weakens encryption's guarantees for everyone.

The Observer's report does not address how the software would coexist with encrypted services, and the campaigners frame the mandate narrowly around content that is already illegal to possess. But a legal requirement to ship classification software on every device sold in a country, and to push it onto older handsets by update, would rank among the most far-reaching technical mandates any government has imposed on consumer hardware. The response from Apple, Google and the wider security community will determine whether Britain becomes the first large-scale test of on-device detection.