Elon Musk’s social media platform X said on Wednesday that it has introduced new safeguards to prevent its AI chatbot, Grok, from generating images that digitally “undress” real people. The move follows widespread criticism over the creation of sexualised images and growing legal and regulatory pressure on Musk’s AI company, xAI.

X said it will now block image-generation features in countries where such content is illegal and prevent users from editing images of real people to make them appear in revealing clothing, such as bikinis or underwear. The restrictions apply to all users, including paid subscribers. As an additional safeguard, image creation through Grok has been limited to paying accounts only.

The announcement comes as California’s attorney general has launched an investigation into whether xAI violated state law by allowing the spread of non-consensual explicit material. Several countries, including Indonesia and Malaysia, have already restricted or blocked access to Grok, while regulators in the European Union, Britain, and India have opened or signalled investigations.