San Francisco (United States) (AFP) – More governments vowed crackdowns Thursday to prevent Grok, the AI chatbot on Elon Musk’s X platform, from undressing or sexualizing images of real people in their jurisdictions, in a face of a growing backlash against the deepfakes. The Philippines became the third country to ban Grok altogether, following Southeast Asian neighbors Malaysia and Indonesia, while Britain and France said they would keep up the pressure.
Several countries have demanded that Musk’s xAI, the developer of Grok, rein in the chatbot after it was used to generate a flood of lewd photos of women and children. X said Wednesday that it would “geoblock the ability” of all Grok and X users to create images of people in “bikinis, underwear, and similar attire” in jurisdictions where such actions are illegal. The announcement came after California’s attorney general launched an investigation into xAI over the sexually explicit material, and several countries either blocked access to Grok or opened their own probes.
“We have implemented technological measures to prevent the Grok account from allowing the editing of images of real people in revealing clothing,” X’s safety team said, adding that the restriction applied to “all users” without exceptions. In an “extra layer of protection,” image creation and the ability to edit photos via X’s Grok account is now available only to paid subscribers, it said.
“I welcome that X is now acting to ensure full compliance with UK law — it must happen immediately,” British Prime Minister Keir Starmer — a favorite target of Musk’s political posts — wrote Thursday on X. “If we need to strengthen existing laws further, we are prepared to do that,” Starmer warned. Meanwhile, Philippines cybercrime chief Renato Paraiso said the country’s block could be effective by the end of the day. He said X’s pledge to limit access would have no effect on the plans, adding that the government would watch to see if the platform follows through on its promises. “We need to clean the internet now because much toxic content is appearing, especially with the advent of AI,” Philippine telecommunications secretary Henry Rhoel Aguda said.
Grok’s so-called “Spicy Mode” allowed users to create deepfakes using simple text prompts such as “put her in a bikini” or “remove her clothes.” An analysis of more than 20,000 Grok-generated images by AI Forensics, a Paris-based nonprofit, found that more than half depicted “individuals in minimal attire” — most of them women, and two percent appearing to be minors. The European Commission, which acts as the EU’s digital watchdog, had said it will “carefully assess” additional measures taken by X to ensure “they effectively protect citizens.”
“France and Europe taking action…is producing results,” Paris’s digital minister Anne Le Henanff told AFP on Thursday, warning that “no platform is above the law.” Her British counterpart Liz Kendall said in a statement that she welcomed X’s move — while backing an investigation by media watchdog Ofcom into whether the images breached British law. California Governor Gavin Newsom had said that xAI’s “vile” decision to allow sexually explicit deepfakes to proliferate prompted him to urge the state’s attorney general, Rob Bonta, to hold the company accountable.
“We have zero tolerance for the AI-based creation and dissemination of nonconsensual intimate images or of child sexual abuse material,” Bonta said Wednesday. He added that the California investigation would determine whether xAI violated state law after the explicit imagery was “used to harass people across the internet.”
Adding further pressure onto Musk’s company Wednesday, a coalition of 28 civil society groups submitted open letters to the CEOs of Apple and Google, urging them to ban Grok and X from their app stores amid the surge in sexualized images. Indonesia on Saturday became the first country to block access to Grok entirely, with Malaysia following the next day. On Thursday, Malaysia’s communications minister said national regulators had found that X’s steps to prevent Grok generating indecent images were “not done in totality.” If X can successfully deactivate and prevent the generation of such online content considered harmful, Malaysia will lift the temporary restriction on Grok, Fahmi Fadzil said.
© 2024 AFP
