Ofcom opens an investigation under the OSA into Grok over the sexualized deepfakes of women and kids being generated on X and threatens X with a ban or a fine
apnews.com/article/grok... Malaysia and Indonesia become the first countries to block Musk's Grok over sexualized AI imagesTruls Gulowsen /@trulsgulowsen:Indonesia, Malaysia, UK, France and others have banned or considered banning X - makes it even stranger for some governments (and media) to maintain it as public information channel — apnews.com/article/grok...Ryan Mac /@rmac:Musk has many supporters who see him as the protector of “free speech.” He's done a lot of personal branding to that
Financial Times
Context & Ripple Effects
The UK action follows fresh scrutiny of X and Grok across Europe, India and Malaysia and India’s order for X to halt obscene Grok outputs or risk its intermediary protections. Indonesia’s temporary block of Grok showed that governments were prepared to restrict access, not merely seek explanations.
Ofcom’s investigation puts X at immediate risk of an OSA fine or service restriction, while making Grok’s image-generation controls and X’s handling of its outputs central to UK regulatory scrutiny.
Users and people depicted in the images gain a formal UK enforcement channel; X must contend with the possibility that failures by its integrated AI tool are treated as a platform-safety issue.
Second-order effects
The UK case adds pressure on X to implement safeguards that work consistently across Grok’s app, website and X, rather than relying on narrower policy changes.
Other authorities assessing the same harms have a concrete precedent for escalating from information requests to access restrictions, as Indonesia’s block of Grok already demonstrated.
Third-order effects
If regulators continue to connect model outputs to the services that distribute them, AI providers and social platforms will face converging obligations around access controls, output moderation and incident response.
The likely longer-run outcome is more jurisdiction-specific availability for high-risk AI features, with service access increasingly shaped by national enforcement choices rather than a single global rollout.
The trend: This is part of the shift toward state-mediated AI access, in which governments use platform-safety rules and service blocks to govern frontier-model deployment.
🧵 We've opened a formal investigation into X under the Online Safety Act, to determine whether it has complied with its duties to protect people in the UK from illegal content. ⬇️ 🔗 https://www.ofcom.org.uk/... [image]
Ofcom will, assuming it follows the template of prior investigations, have sent X a letter warning it of that investigation, threatening X's senior managers with arrest if they refuse to comply, and demanding X produce information without first getting a warrant.
NEW: @Ofcom has launches a formal investigation into X over the use of the Grok to manipulate images of women and remove their clothes. “We have decided to open a formal investigation to establish whether X has failed to comply with its legal obligations under the Online Safety
(APNews) While Musk is not taking any steps to prevent his AI Grok from creating oversexualized AI Images, countries are beginning to block his AI engine. Stay tuned. — apnews.com/article/grok... Malaysia and Indonesia become the first countries to block Musk's Grok over sexu…
Indonesia, Malaysia, UK, France and others have banned or considered banning X - makes it even stranger for some governments (and media) to maintain it as public information channel — apnews.com/article/grok...
Musk has many supporters who see him as the protector of “free speech.” He's done a lot of personal branding to that effect. But I suspect some of them would be uncomfortable with “free speech” encompassing the ability to create & share nonconsensual sexual material. — www.ny…
Great to see Ofcom launch a formal investigation into X. It's crucial this doesn't stop at AI imagery, it must also cover the vast amount of other illegal content X isn't removing under the Online Safety Act — https://tinyurl.com/mr2syd99
www.ofcom.org.uk/online-safet... Good. Let's hope it goes somewhere and doesn't fade to nothing after any, er, persuasion from very very rich people...
We are pleased to see this from Ofcom www.ofcom.org.uk/online-safet.... However, it appears to have a main or exclusive focus on CSAM and there is a wide variety of other illegal content which X, in breach of its duties under the Online Safety Act, has failed to remove.
There is nothing wrong per se with Ofcom's statement here - they're following the process they have to as soon as possible - but it does show how slow and bureaucratic state processes are compared to how the internet moves. — Makes it a far less effective regulator
UK communications regulator has announced it's opening an official investigation into X over non-consensual imagery. This follows it getting in touch with X last week. — It's investigating on a host of grounds: from whether X has done proper risk assessments to failing to take…
We know you're reading, Apple. For maybe the first time in the '20s, do the right thing by users instead of your shareholders and block the child porn factory that is X and Grok until they make changes. — https://appleinsider.com/...