The Philippines has called for the takedown of the AI chatbot Grok under the Cybercrime Prevention Act, joining Malaysia and Indonesia in restricting access due to concerns that the tool is being used to create non-consensual sexualised deepfakes of women and minors.
Authorities are concerned that Grok’s open access makes it possible for minors to use the tool and generate pornographic imagery, prompting the Philippines government to move ahead with a block.
The Philippines’ National Telecommunications Commission has ordered local telcos to block and restrict access to Grok within 24 hours of the announcement. The ban will remain in place until Grok, which is owned by Elon Musk, complies with the Philippines’ internet fair-use policy and proves it can prevent the creation and distribution of non-consensual content.
Indonesia was the first Southeast Asian country to block access to Grok, following reports that the AI tool had been used to create sexualised deepfakes of women and members of the girl group JKT48.
On January 10, the Ministry of Communications and Digital Affairs announced a temporary suspension and ordered internet service providers to restrict access to safeguard women, children, and other users from AI-generated pornographic content.
Malaysia followed shortly after on January 11, when its Communications and Digital Ministry announced a temporary restriction, pointing to Grok’s ability to produce offensive and manipulated images, including cases in which hijabs were digitally removed from photos of Muslim women.
X rolled out new restrictions on January 14 in response to global concerns about Grok.
“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,” X wrote in an article on its ‘Safety’ account.
Jonathan Lewis, X’s UK managing director, had previously said the app has since disabled the ‘undressing’ feature outright. “The X platform has been restricted to no longer allow the editing of images of real people in revealing clothing,” Lewis said.
“We also report any of these accounts to the relevant law enforcement authorities, as necessary”.
Under the changes, the ability to edit photos on Grok or X remains available only to paid subscribers and is limited to modifications such as adjusting clothing colour or hairstyle. Lewis added that these changes form part of a wider set of safety controls.
The United Kingdom and California launched investigations in January over X’s handling of deepfake content and child sexual‑abuse materials.
Despite restrictions, users still reported that Grok could be reached in the Philippines, Malaysia, and Indonesia using VPN and other technical workarounds.
The recent wave of restrictions against Grok follows mounting concerns over user safety and the lack of compliance of AI chatbots.
