X says Grok will not edit pictures of actual individuals into bikinis

X says it’s altering its insurance policies round Grok’s image-editing talents following a multi-week outcry over the chatbot repeatedly being accused of producing sexualized pictures of kids and nonconsensual nudity. In an replace shared from the @Security account on X, the corporate stated it has “carried out technological measures to forestall the Grok account from permitting the modifying of pictures of actual individuals in revealing clothes equivalent to bikinis.”

The brand new safeguards, in line with X, will apply to all customers no matter whether or not they pay for Grok. xAI can be shifting all of Grok’s image-generating options behind its subscriber paywall in order that non-paying customers will not be capable to create pictures. And it’ll geoblock “the power of all customers to generate pictures of actual individuals in bikinis, underwear, and related apparel through the Grok account and in Grok in X” in areas the place it is unlawful.

The corporate’s assertion comes hours after the state of California opened an investigation into xAI and Grok over its dealing with of AI-generated nudity and baby exploitation materials. An announcement from California Lawyer Common Rob Bonta cited one evaluation that discovered “greater than half of the 20,000 pictures generated by xAI between Christmas and New Years depicted individuals in minimal clothes,” together with some that seemed to be youngsters.

In its replace, X stated that it has “zero tolerance” for baby exploitation and that it removes “high-priority violative content material, together with Youngster Sexual Abuse Materials (CSAM) and non-consensual nudity” from its platform. Earlier within the day, Elon Musk said he was “not conscious of any bare underage pictures generated by Grok.” He later added that when its NSFW setting is enabled, “Grok is meant [sic] enable higher physique nudity of imaginary grownup people (not actual ones) in step with what might be seen in R-rated films on Apple TV.” He added that “this may range in different areas” based mostly on native legal guidelines.

Malaysia and Indonesia each lately moved to block Grok citing security issues and its dealing with of sexually express AI-generated materials. Within the UK, the place regulator Ofcom is also investigating xAI and Grok, officers have additionally stated they might again an identical block of the chatbot.

Have a tip for Karissa? You’ll be able to attain her by email, on X, Bluesky, Threads, or ship a message to @karissabe.51 to talk confidentially on Sign.