Malaysia and Indonesia have blocked access to Grok, an artificial intelligence tool integrated into X, after authorities said it was being used to generate harmful deepfake images. Officials in both countries said the action was taken to limit the spread of non-consensual and sexualised content created using the tool.
In Indonesia, the Ministry of Communication and Informatics said Grok was suspended after it was found to be capable of producing manipulated images that violate national laws and social norms. The ministry said existing regulations allow authorities to restrict digital services that enable the distribution of prohibited content, including material that harms individuals or undermines public order. Officials said the block will remain in place until safeguards are strengthened.
Malaysia’s communications regulator took similar action, citing concerns that Grok’s image generation features could be misused to create deepfakes without adequate controls. The regulator said the platform did not demonstrate sufficient technical measures to prevent abuse and relied too heavily on user reporting rather than proactive moderation. Access to the tool was restricted while authorities assessed compliance with local content and safety rules.
Grok is developed by xAI, which is linked to X through common ownership. The tool gained attention after users demonstrated its ability to generate altered and sexualised images from text prompts. Critics said such capabilities pose risks to privacy and personal safety, particularly for women and children.
Both governments said the restrictions are precautionary and focused on user protection. Indonesian officials said they have requested clarification from X and xAI about how the tool is being used locally and what measures are in place to prevent misuse. Malaysian authorities said access could be restored if the operators demonstrate effective controls that meet regulatory standards.
The blocks mark the first known national-level restrictions on Grok linked specifically to deepfake concerns. The moves follow wider international scrutiny of generative artificial intelligence tools and their potential to create harmful or misleading content. Regulators in several countries have said existing laws may apply to AI systems if they enable illegal or damaging material.
xAI has previously limited some image generation features and adjusted access settings following criticism. Authorities in Malaysia and Indonesia said further technical and policy changes would be required before the service could resume normal operation.