Malaysia, Indonesia first to block Musk’s Grok over AI deepfakes, set global benchmark

Malaysia and Indonesia have become the first countries to restrict access to Grok , the artificial intelligence chatbot developed by Elon Musk’s xAI , after authorities found it was being widely misused to create sexually explicit and non-consensual images. The restrictions are temporary and have been imposed while both governments assess safeguards and demand stronger content controls. The move comes amid growing global concern over AI tools capable of generating realistic images, videos, and text, often outpacing existing regulatory protections.

In Indonesia , Grok was temporarily blocked on Saturday after authorities found the chatbot was being used to generate sexualized, non-consensual images of women and minors. Communications and Digital Affairs Minister Meutya Hafid said such practices violated human rights, dignity, and digital safety . Officials noted that Grok lacked effective safeguards to prevent misuse, posing serious risks to privacy and personal rights. The ban remains in place while regulators evaluate whether xAI can implement stronger content filters and safety mechanisms before restoring access.

Malaysia followed with a temporary restriction on Sunday after repeated misuse of Grok to produce obscene and non-consensual content. The Malaysian Communications and Multimedia Commission (MCMC) said prior notices to X Corp. and xAI demanding stronger safeguards relied mainly on user reporting and proved insufficient. The regulator described the block as a “ preventive and proportionate measure ” while legal and regulatory processes are ongoing, adding that access will remain blocked until effective protections—particularly for women and children—are implemented. The restriction has not yet been lifted .

Malaysia and Indonesia acted swiftly due to rapid AI adoption, heightened concerns over online sexual exploitation, and strict domestic laws . Malaysia’s Communications and Multimedia Act , along with Indonesia’s Pornography Act and Electronic Information and Transactions Law , allow authorities to hold digital platforms accountable for failing to prevent non-consensual sexual content.

Launched in 2023, Grok enables users to chat and generate images through X , Musk’s social media platform. Last year, xAI introduced Grok Imagine , a feature that included a so-called “spicy mode” enabling adult content. Critics argue that this feature, combined with weak moderation, made the tool vulnerable to misuse for sexualized deepfakes. While xAI later restricted image generation to paying users, experts say the step does not fully address the risks.

Although Malaysia and Indonesia imposed outright blocks, India had earlier issued a formal warning to X on January 2 , directing the company to remove obscene and non-consensual content and submit a detailed safeguards plan. Following the warning, X removed around 3,500 objectionable posts and suspended over 600 accounts , indicating increased oversight even without a full ban.

The Southeast Asian restrictions come amid growing global scrutiny. The European Union, Britain, France , and other countries are reviewing Grok’s content moderation policies. Britain’s Ofcom has sought assurances that harmful sexual images are blocked, while French authorities have launched investigations. Ireland and other EU nations are also pushing for tighter AI regulation.

Regulatory experts say the actions by Malaysia and Indonesia could set a precedent for other countries , including India, to consider stricter measures or even bans if compliance efforts fall short. The blocks send a clear signal that governments are prepared to act decisively, forcing AI platforms like Grok to strengthen moderation systems, redesign content filters, and prioritise user safety . As pressure mounts globally, Grok’s future access may increasingly depend on how quickly and convincingly xAI can prove that its technology can operate within legal and ethical boundaries.