Malaysia suspends access to Musk's Grok AI

Malaysia suspended access to Elon Musk's chatbot Grok over AI-generated pornographic content, the country's tech regulator said on Sunday.

The decision follows global backlash after it emerged that Grok's image creation feature allowed users to sexualise pictures of women and children using simple text prompts.

On Saturday Indonesia became the first country to deny all access to the tool, which has been restricted to paying subscribers elsewhere.

The Malaysian Communications and Multimedia Commission said in a statement it had "directed a temporary restriction on access to the Grok artificial intelligence for users in Malaysia" with immediate effect.

When an AFP reporter in the Malaysian capital Kuala Lumpur fed Grok prompts on Sunday, there was no response.

"This action follows repeated misuse of Grok to generate obscene, sexually explicit, indecent, grossly offensive and non-consensual manipulated images," the regulator said.

The statement cited "content involving women and minors, despite prior regulatory engagement and formal notices" issued to Musk's X Corp. and xAI startup which developed Grok.

The AI tool is integrated into social media platform X.

The Malaysian regulator said it deemed the platform's safeguards inadequate, adding that access would resume only after the required changes are verified.

X Corp. had "failed to address the inherent risks posed by the design and operation of the AI tool", relying "primarily on user-initiated reporting mechanisms", the regulator said.

European officials and tech campaigners on Friday slammed Grok after its controversial image creation feature was restricted to paying subscribers, saying the change failed to address concerns about sexualised deepfakes.

Grok had appeared to deflect the criticism with a new monetisation policy, posting on X on Thursday that image generation and editing were now "limited to paying subscribers", alongside a link to a premium subscription.