The Guardian view on children and the internet: rolling back big tech’s untrammelled power | Editorial

A belated change of policy on nude digital images of children must be part of a wider reset

A mid the flurry of resignations by ministers who said they had lost confidence in Sir Keir Starmer’s leadership, Jess Phillips’s attack on his record on tech regulation stood out. “Over a year ago I presented solutions, long worked on by brilliant civil servants, that would end the ability for children in the UK to take naked images of themselves,” she wrote. The postponement of an announcement in March left her frustrated. In the end, all that Ms Phillips managed to secure was a pledge that the law might change sometime.

Other campaigners echoed her frustration. Hannah Swirsky, head of policy at the Internet Watch Foundation, agreed that the government had been slow to act, despite the rise in offences involving self‑generated explicit imagery.

One month on, Ms Phillips’s decision to quit her role as safeguarding minister has perhaps focused minds. Sir Keir started the week with a speech at London Tech Week in which he announced that Google and Apple – which between them control the operating systems on almost all smartphones – have until September to install software that blocks all nude images from children’s phones. He highlighted a product built by a UK company, SafeToNet, that shows it can be done.

Technically, the announcement was only a warning. The deadline for action is three months off and the Online Safety Act was years in the drafting. But it is part of a broader shift. For years, tech companies have fought plans to regulate their systems, preferring rules to focus on users – so that individuals rather than platforms are held responsible for any harms. Some digital rights activists have joined this resistance. They oppose identity- and age‑verification tools on principle, despite the obvious risks to children in an engineered environment that refuses to recognise their age.

But in recent months legislators and courts have pushed back. Last December, the world’s first social media ban for under-16s came into force in Australia. In March, a court in California ruled in favour of a woman who claimed that Meta and YouTube were responsible for her social-media addiction as a child. Now, the UK government is demanding that tech businesses go beyond their content-moderation duties, by altering settings on their devices – with further announcements on age limits expected soon. The children’s commissioner for England, Rachel de Souza, believes that new rules should also apply to 16- and 17-year-olds. They too are children under the terms of the UN Convention on the Rights of the Child .

The Online Safety Act is meant to protect adults as well as children, as is the EU’s Digital Services Act . Belatedly, some legislators appear to be realising that they need to get better at anticipating damage, rather than responding to it. This can be harm to individuals – as in the case of Jess Asato, the MP who is suing xAI over the use of Grok to produce fake, sexualised images of her – or to society as a whole. A new report from the Social Market Foundation thinktank describes local misinformation as a “silent killer of trust”.

Firm action from governments on child safeguarding is overdue; the laissez-faire approach of the last two decades has been a shameful failure. But other changes will be needed, too, if big tech is to be cut down to size as a tool used by humans – rather than people being treated as pawns by big tech.