TL;DR
The Advertising Standards Authority has banned five adverts for AI companion and video-generation apps that ran on Meta platforms, ruling they objectified and sexualised women and traded in harmful gender stereotypes. One depicted a minor in sexualised terms. The regulator found them with an AI monitoring tool of its own, and not one of the developers answered its enquiries.
What was banned
The adverts promoted tools for turning photographs of real people into sexual imagery. Two for a portrait generator called Nexaipic urged viewers to supply a picture of someone they fancied and let the software do the rest. An image-to-video app listed only as KH31 DD22 showed a clothed woman becoming topless. PictoPop used a submissive pose to demonstrate its output, and Rusto AI advertised sexualised video made from ordinary photos.
The fifth, from Animcha Ltd, promoted an AI companion: a synthetic female character presented as customisable and instructable, pairing sexual content with props and clothing that suggested a child. The ASA ruled it portrayed a minor sexually.
The ASA’s Nicky Baker said the regulator takes “a firm line on ads that are likely to cause serious or widespread offence”. Encouraging “the objectification of women and the creation of explicit content from people’s images”, Baker added, carries “the potential to cause real-world harm”.
The regulator is using the technology it is policing
The ads surfaced through the watchdog’s “Active Ad Monitoring System”, which hunts proactively for likely breaches rather than waiting on complaints. That deserves noting on its own terms: a UK regulator running machine-scale detection against machine-scale advertising, with no new legislation needed to do it.
The legal backdrop shifted mid-investigation. Britain outlawed apps producing non-consensual nude imagery earlier this year, following anger at Grok, the chatbot on X, for generating nude depictions of women on request. Future cases, the ASA says, will be measured against statute as well as its own code, and it may refer illegal material to platforms and police.
Meta’s numbers tell their own story
Meta had pulled the five ads under its own rules before the ASA acted, and declined to comment. Its disclosed figures show the scale involved: upwards of 344,000 nudification-app adverts removed across Facebook and Instagram in the three months to January 2026. Researcher Alexios Mantzarlis counted 11,336 live ads on Meta platforms in early September claiming to strip clothed subjects.
Looking forward
Enforcement here is fast but shallow. The ASA can stop an advert; it cannot reach the app, and the companies did not even reply. For UK firms, the practical read is that advertising rules are currently the quickest-moving lever on generative AI products, well ahead of any statutory AI framework.