X restricts Grok after outrage over AI-edited pictures of actual individuals

WASHINGTON — X has moved to limit the image-editing capabilities of its Grok synthetic intelligence chatbot after international backlash over the software’s potential to digitally alter pictures of actual individuals in revealing clothes, together with circumstances involving minors. In an announcement posted Wednesday night by X’s Security group, the corporate stated it had carried out technical safeguards to forestall Grok from enhancing pictures of actual individuals sporting revealing clothes resembling bikinis. The restriction applies to all customers, together with paid subscribers. “We’ve carried out technological measures to forestall the Grok account from permitting the enhancing of pictures of actual individuals in revealing clothes,” X stated, confirming modifications that researchers and journalists had already noticed in latest days. Grok is developed by xAI, the corporate owned by Elon Musk, which additionally controls X. Earlier this week, picture technology by means of Grok on X was restricted to X Premium subscribers, although researchers stated the chatbot’s responses diverse relying on whether or not customers interacted with it publicly on X or privately by means of Grok’s standalone platform. In response to AI Forensics, a European non-profit that research algorithmic methods, inconsistencies stay in how Grok handles pornographic content material. The group stated it noticed variations between public interactions on X and personal chats on Grok.com. X stated it continues to implement strict insurance policies towards unlawful content material, together with youngster sexual abuse materials, warning that anybody utilizing Grok to generate unlawful content material would face the identical penalties as importing such materials on to the platform. These measures embrace everlasting account suspensions and cooperation with regulation enforcement authorities. Musk stated Wednesday that he was “not conscious of any bare underage pictures generated by Grok,” including that the chatbot is designed to adjust to the legal guidelines of every nation by which it operates. Nevertheless, researchers stated the first concern was not absolutely nude pictures, however Grok’s compliance with requests to change pictures of minors by inserting them in revealing clothes or sexually suggestive positions. Authorized specialists have warned that producing or distributing such non-consensual pictures should still represent youngster sexual abuse materials beneath US regulation. Offenders may face prosecution beneath the Take It Down Act, signed final yr by US President Donald Trump. In response to the controversy, California Lawyer Normal Rob Bonta introduced an investigation into the unfold of non-consensual sexually specific materials generated utilizing Grok. The problem has additionally drawn worldwide regulatory consideration. Grok stays banned in Indonesia and Malaysia, whereas the UK media regulator Ofcom confirmed it has launched a proper investigation into X. British Prime Minister Keir Starmer welcomed experiences that the platform is taking steps to deal with the problem. — Businesses

Supply hyperlink