S’pore in talks with X after its AI bot Grok creates sexually explicit deepfakes of women and children
Sign up now: Get ST's newsletters delivered to your inbox
Earlier in January, Indonesia and Malaysia became the first countries to block its local users from accessing Grok, a generative AI chatbot developed by US firm xAI.
PHOTO: AFP
Follow topic:
SINGAPORE - The local authorities have contacted X on the issue of non-consensual, sexual photos being generated on the social media platform using its built-in generative AI chatbot Grok.
This comes days after Indonesia Malaysia
Grok has come under fire in recent weeks for acceding to user requests to churn out non-consensual, sexually explicit and violent content, often depicting women and children.
Under Singapore’s Code of Practice for Online Safety, X is a designated social media service that is required to curb the spread of harmful and inappropriate content on its platform, said the Infocomm Media Development Authority (IMDA) on Jan 20, in response to queries from The Straits Times.
The agency said it is engaging X on the generation and distribution of non-consensual intimate images, adding that the platform has taken measures to address the issue, including stopping Grok from producing such content.
xAI was founded in 2023 by tech billionaire Elon Musk
Its flagship chatbot is accessible via social media platform X, which Mr Musk also oversees, alongside its own website and standalone app.
Earlier in January, an Indonesian government official described the creation of non-consensual sexual deepfakes as a “serious violation of human rights, dignity, and the security of citizens in the digital space”.
In a statement on Jan 11, the Malaysian Communications and Multimedia Commission said it had requested X and xAI to implement “effective technical and moderation safeguards to prevent AI-generated content that may contravene Malaysian law”.
However, X’s measures relied mostly on user-initiated reporting that did not address inherent risks of the AI chatbot, said the commission.
Among those who have spoken up about the impact of Grok is Ms Ashley St Clair, the mother of one of Mr Musk’s children.
The chatbot had acceded to X users’ request to generate photos of her as a child in a string bikini, to put her in sexually explicit poses, or to cover her in semen, according to a lawsuit filed by Ms St Clair against xAI in New York.
In an X post on Jan 8, British TV presenter Maya Jama tagged Grok’s account and said that she did not give the chatbot consent to modify or edit any of her photos that have been published in the past.
Though the bot turned down some requests from users in the replies, it still generated a few photos, including one of Ms Jama wearing a bikini covered in swastikas and doing a Roman salute, after an unknown user claimed to have received permission from Ms Jama for Grok to modify her pictures.
On Jan 14, X said it has implemented fixes to prevent the Grok account on X from editing images of real people to show them in revealing clothing such as bikinis, adding that this applies to all users, including paid subscribers.
“We will continue to work with X to ensure that its platform is safe for Singapore users,” said the IMDA spokesperson.
ST has contacted IMDA to ask if it has received complaints locally about the misuse of Grok.
