We have implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing such as bikinis. This restriction applies to all users, including paid subscribers, reads an announcement on X, which operates the Grok AI tool.
The change was announced shortly after California's top prosecutor indicated that the state was examining the proliferation of sexualised AI deepfakes, including those of minors, produced by the AI model.
They also confirmed that only paid users would have editing access using Grok on its platform, which aims to enhance accountability for misuse.
With NSFW settings enabled, Grok is designed to allow upper body nudity of imaginary adult humans (not real ones) consistent with rates allowable in R-rated movies, as Musk explained. This standard will vary by country according to national laws.
Global leaders have voiced concerns over Grok's editing capabilities, leading to its ban in Malaysia and Indonesia after reports of unauthorized alterations resulting in explicit imagery.
Ofcom, the media regulator in Britain, has commenced investigations into whether X has violated UK laws regarding sexual images. Following the outcry, some UK MPs decided to leave the X social media platform.
California Attorney General Rob Bonta remarked that this type of explicit content, showing individuals in compromising situations, is being used for harassment online.
Amidst the ongoing criticism, policy researcher Riana Pfefferkorn highlighted that the implementation of these safeguards was overdue and stressed the need for clarity on enforcement mechanisms against rule violations.
