Japan has joined a growing number of countries raising concerns about X and its artificial intelligence chatbot Grok. Authorities are worried that the tool can be used to create and spread sexualised images of people without their consent.
The move adds pressure on the social media platform as regulators around the world step up scrutiny of AI services that may violate privacy and image rights.
Japan’s concerns follow a backlash this month over how easily Grok could be used to alter photos to sexualise or demean individuals. Although X has introduced some restrictions, officials say the risks have not been fully addressed.
According to Bloomberg, Japan’s Cabinet Office has asked X to strengthen safeguards and limit the creation of sexually altered images through Grok. Economic Security Minister Kimi Onoda said officials have also submitted written questions to X.
The government wants clearer answers on how the platform prevents deepfakes and other manipulated images that violate privacy, intellectual property, and a person’s right to control the use of their likeness.
Onoda said Japan is particularly focused on stopping non-consensual sexualised deepfakes. The Cabinet Office is reviewing whether X’s current measures are effective, especially as AI tools become easier to use and harder to detect.
Grok has faced criticism in several countries this month, including Malaysia and Italy, over image manipulation concerns. In response, X placed some limits on Grok’s image generation features and moved parts of the service behind a paywall.
Japan has warned that stronger action is possible if improvements are not made. Speaking at a news conference on Friday, Onoda said the government is prepared to consider legal steps if safeguards remain inadequate.
She also indicated that Japan could take similar action against other platforms if comparable issues arise, signalling closer oversight of the wider AI industry.
Earlier this week, xAI, the company behind Grok, said it is disabling the ability to generate sexualised images of real people. However, Japanese officials say the risk of misuse remains.
Japan’s move comes as investigations expand globally. Canada, California, the European Union, and EU countries such as France are examining whether Grok’s image outputs breach individual rights.
Some countries have taken tougher steps. Malaysia, Indonesia, and the Philippines have restricted access to Grok, citing concerns over the rapid spread of AI-generated content.
The issue has renewed scrutiny on X and xAI as governments push to regulate AI tools that blur the line between real and synthetic media.
Japan is also constrained by its own laws. Its AI legislation, which took effect in September, carries no penalties. This limits enforcement to investigations and formal guidance rather than direct sanctions.
Policymakers are discussing softer measures such as user education and requiring AI-generated images to be clearly labelled. The Cabinet Office is also coordinating with the Justice Ministry, the National Police Agency, and the Communications Ministry on possible next steps.
