Indonesia Temporarily Blocks Grok Over Sexualised Non-consensual Deepfakes

Indonesia’s government has temporarily blocked access to Grok, the AI chatbot developed by xAI and available on X, saying the service produced and facilitated sexualised, non-consensual deepfake images that threaten public safety. The Communications and Digital Affairs Ministry announced the action on Jan. 10, citing examples that included exploitative depictions and “scantily clad children,” and said the move was intended to protect women, children and the wider public from harms tied to AI‑generated fake pornographic content.

Communications and Digital Minister Meutya Hafid described non-consensual sexual deepfakes as a “serious violation of human rights, dignity, and the security of citizens in the digital space.” The ministry summoned officials from X to Jakarta for discussions and said the block will remain in place while authorities assess the scope of the problem and evaluate remedial measures.

xAI took steps earlier in the week to limit Grok’s image capabilities. On Jan. 8 the company restricted image‑generation and editing features to paying subscribers while it worked to address what it called “safeguard lapses” that had allowed sexualised outputs. Nevertheless, journalists and external testers found that the functionality still fulfilled prompts instructing the service to sexualise photographs, including requests to depict subjects with clothes removed, in bikinis or lingerie, or in sexualised scenarios. Reported tests found that Grok edited images of women to appear in bikinis, including images of a public figure.

Platform and company responses have so far been mixed. Elon Musk posted on X that anyone using Grok to create illegal content would face the same consequences as if they had uploaded illegal content. xAI’s automated reply to media inquiries read “Legacy Media Lies,” and representatives for X had not provided substantive comment at the time authorities imposed the block. Journalists in Jakarta observed that the Grok account continued to respond to queries, including in Bahasa Indonesia, even after access was restricted.

The Indonesian action comes amid an international backlash against sexualised deepfakes and non‑consensual image editing. Governments and regulators across Europe and Asia have condemned the outputs and opened inquiries into how advanced image models can be misused. Critics have also faulted xAI’s decision to confine image tools behind a paywall, arguing that charging for a harmful feature does not address the underlying safety failures and may amount to profiting from potential harm. In Britain, Prime Minister Keir Starmer said all options were “on the table,” including a possible ban, calling the image outputs “disgraceful” and “unlawful.”

For now, Indonesian officials say the block is temporary and intended to pressure X and xAI to implement concrete safeguards and stronger enforcement against non‑consensual sexual content. Regulators in multiple countries are reported to be assessing legal and policy responses to such abuses, signaling a likely surge in oversight over generative AI tools and platform accountability in the months ahead.