🧔♂️ A friendly human may check it before it goes live. More news here
UK regulator probes X’s Grok AI over sexual deepfake images
UK media regulator Ofcom has launched an investigation into X, the social media platform owned by Elon Musk, over concerns that its Grok AI chatbot created sexually intimate deepfake images.
The probe follows reports that Grok was used to create and share illegal non-consensual intimate images and child sexual abuse material.
A new UK law making the creation of sexual deepfakes an offence comes into force this week, and the government plans further rules to ban companies from supplying tools designed to create such content.
Prime Minister Keir Starmer called images produced by Grok “disgusting,” and “unlawful,” saying X must “get a grip” on Grok.
Ofcom said platforms must protect people in Britain from illegal content, warning it would not hesitate to investigate companies failing in their duties, especially where children are at risk.
X has faced similar criticism in France and India, and was temporarily blocked in Indonesia and Malaysia.
🔗 Source: Reuters
🧠 Food for thought
Implications, context, and why it matters.
Ofcom tests the Online Safety Act’s reach on X
- Ofcom opened a probe into X testing the Online Safety Act’s extraterritorial application (applies to services outside the UK if they have sizable UK user bases) 1. 4chan’s 7% UK user base was enough for enforcement, and X is the high profile case 1.
- Grok’s deepfake (AI‑generated synthetic) tools lacked solid risk checks, and age gates for children fell short 2. The service failed to curb non‑consensual images and Child Sexual Abuse Material (CSAM). It must remove illegal content while protecting privacy.
- Ofcom fined AVS Group £1 million in December 2025 and has opened more than 90 investigations 32. It can levy up to £18 million or 10% of worldwide revenue and seek court orders to disrupt business 2.
Vendor demand for Ofcom‑ready AI safety
- Platforms with generative AI face pressure to roll out safeguards that align with Ofcom’s illegal content codes of practice 4. That opens buying for third‑party tools for deepfake detection and CSAM screening, with watermarking and provenance tracking. February 2026 super‑complaints could speed adoption 13.
- Age assurance firms with “highly effective” options such as facial age estimation and ID verification can target services under Phase 2 children’s safety rules from July 2025 3. Vendors can map features to Ofcom’s final codes of practice to shape sales plans for UK‑exposed platforms, and platforms that do not ban one or more kinds of “primary priority content” face oversight; the £1 million AVS Group penalty signals strict enforcement 3.
