Elon Musk’s artificial intelligence company xAI has announced new limits on its chatbot Grok after strong criticism from governments and the public due to an “undressing spree”. The company said Grok will no longer be allowed to edit photos of real people to show them in revealing clothes like bikinis or underwear.
Elon Musk said he was “unaware” of Grok creating sexual images of real people. (AFP)
These new rules apply to everyone on X, including people who pay for premium access, news agency Bloomberg reported. xAI said the step was taken after many users misused Grok to create fake sexual images of women and children without their permission.
ALSO READ | Has Elon Musk filed for one-year-old son’s full custody from Ashley? All on billionaire’s latest move
Over the past few weeks, thousands of such images were shared on X, causing anger and concern around the world, said the report.
Musk’s stance
Elon Musk has defended Grok, saying the tool only responds to user requests and is designed to refuse illegal content. Musk also said he was “not aware” of any naked images of underage children created by Grok, a Reuters report noted.
A formal investigation underway
California’s Attorney General Rob Bonta has now launched a formal investigation into xAI to see if the company broke the law, according to an ANI report. He said reports of non-consensual sexual images made using Grok were “shocking” and must be stopped immediately.
Bonta also stressed that there is zero tolerance for content that harms children or creates sexual images without consent.
ALSO READ | Elon Musk calls medical school ‘pointless’ as he predicts AI will outperform human doctors
California Governor Gavin Newsom supported the move and asked officials to hold xAI accountable, as per the Bloomberg report.
How the world responded to AI deepfake abuse
Other countries are also taking action, with investigations opened in the UK and France, and checks by the European Union.
ALSO READ | Elon Musk, X warn of ‘consequences’ after uproar over Grok ‘undressing’ spree
Indonesia has fully blocked access to Grok, while Malaysia has restricted it due to similar concerns, the ANI report said. In India, the IT ministry has written to X asking what steps are being taken to stop obscene AI content.
(With inputs from Bloomberg, Reuters and ANI)