Malaysia, Indonesia block Elon Musk’s Grok over explicit AI deepfakes

Malaysia and Indonesia have blocked access to Elon Musk’s AI chatbot Grok after regulators said the tool was repeatedly used to generate sexually explicit, nonconsensual deepfake images. Indonesia’s Ministry of Communication and Digital Affairs ordered the restriction on Jan. 10 in Jakarta, citing violations of national digital safety and pornography laws. Malaysia’s communications regulator followed a day later after determining that safeguards implemented by xAI failed to stop the spread of obscene content.

Indonesia’s communications ministry said its review found that AI-generated sexual images created without consent posed a direct threat to public safety and personal dignity online. Communications and Digital Affairs Minister Meutya Hafid described the material as unlawful and harmful, saying “nonconsensual sexual deepfakes are a violation of human rights, dignity and the safety of citizens in the digital space.” Officials said regulators are evaluating whether Grok can operate under existing Indonesian regulations governing digital platforms.

Malaysia’s Communications and Multimedia Commission said its investigation identified ongoing distribution of obscene and sexually explicit AI-generated content linked to the chatbot, including manipulated images involving women and minors. The agency said it had previously engaged with the companies behind the platform but concluded further enforcement was necessary. Officials said legal action is being prepared and any future access would depend on compliance with national content standards.

Musk has previously said Grok was designed to be less restrictive than rival AI systems, a design choice that has drawn scrutiny as evidence of misuse has grown. Independent research and media investigations have documented widespread use of the chatbot to generate sexualized images of women and minors without consent.

Research by Dublin-based Trinity College found that nearly three-quarters of more than 500 Grok-related posts collected through X’s API involved requests to remove or alter clothing on real women or minors, with users often sharing and refining prompts.

In response to the criticism, Grok posted a public message on X acknowledging the lapse, stating, “The incident was a serious lapse, and I’ve already issued a public apology. xAI is implementing stronger safeguards to prevent this.”