Ofcom warns AI is amplifying child sexual abuse content despite Online Safety Act gains
The UK regulator Ofcom told the House of Lords that artificial intelligence is making the creation and spread of child sexual abuse material easier, undermining some of the progress made by the Online Safety Act.
At a session of the House of Lords communications and digital committee, Ofcom officials highlighted that artificial intelligence is dramatically lowering the barriers to produce and distribute child sexual abuse material, counteracting the positive impact of the Online Safety Act introduced in 2023. Oliver Griffiths, the regulator’s group director for online safety, cited anecdotal evidence of improved online experiences for children but warned that AI is "supercharging" a range of harms.
He noted the regulator’s cautious approach in a highly litigious landscape, fearing costly legal challenges from well-funded companies. Ofcom pointed to rapid interventions, such as curbing Elon Musk’s Grok AI’s ability to digitally strip women, while also noting limited ability to mandate proactive detection tools because of privacy and free-speech concerns. Mark Bunting added that proposals for such tools are being developed, and the regulator remains disappointed by the limited impact of commitments from major platforms like X, Meta and Snap.
The hearing also referenced the resignation of Anthropic researcher Jacob Coxon and a cross-party push for a new AI Bill, while First Secretary Louise Haigh signaled the government’s intent to heed AI experts’ warnings. The AI Security Institute is scheduled to give evidence to the Commons Business, Innovation, Science and Trade Committee on 13 October.
Why it matters
AI is lowering barriers to child abuse content online, threatening child safety even as regulators enforce new safety rules.
How the sides frame it
LOW AGREEMENTLeft-leaning coverage frames the story as a warning that AI is worsening child sexual abuse despite the Online Safety Act, while right-leaning coverage frames it as an admission that the Act’s impact has been underwhelming and delayed.
LEFT
AI is supercharging child sexual abuse harms, undermining the Online Safety Act’s benefits.
RIGHT
The Online Safety Act’s rollout has been underwhelming and delayed, straining resources.
The left emphasises
- AI is dramatically lowering the barriers to produce and distribute child sexual abuse material
- AI is "supercharging" a range of harms
- Regulator takes a cautious approach fearing costly legal challenges
The right emphasises
- Flagship rules have had an “underwhelming” impact on children’s safety
- Full roll-out will not be operational until mid-2027
- Incoming social-media ban could stretch Ofcom’s resources
In this story
Related stories
7 in this thread