Japan joins long list of countries investigating Musk’s Grok

On Friday, Japan announced it is investigating X due to the inappropriate images generated by Musk’s AI service, Grok. The government plans to explore all measures to stop inappropriate images from being created.

Economic Security Minister Kimi Onoda stated that Japan’s Cabinet Office asked X to make immediate improvements, but the company has not responded yet.

Tokyo pressures X for answers about Grok’s controls

Tokyo officials sent written questions to X about steps to stop deepfakes and images that breach privacy, intellectual property, and likeness rights.

Onoda, who is also a minister of state for AI strategy, said at a Friday news conference that all options, including legal action, are being considered if there is no improvement. She added that they will respond if similar issues arise on other social media platforms.

The minister said it is troubling that people can still generate such images. However, she noted that AI itself is not to blame. She believes the Japanese government must explore various solutions to the true problem.

Onoda spoke after Britain and Canada announced they are advancing their investigations into Grok.

Japan seeks to regulate AI while aiming to rival the U.S. and China in this vital field. Its AI law, effective since September, has no penalties and restricts government action to investigations and official guidance for violations.

Authorities worldwide move against Grok

Authorities worldwide have sought to restrict the AI chatbot after users discovered it could produce sexualized images of women and minors.

This month, Grok faced heavy criticism from users and governments in countries from Malaysia to Italy for how easily it could be used to alter photos in a sexual or degrading way

The Philippines and Indonesia have temporarily banned access to Grok due to the generation of explicit images.

Canada, California, the EU, and EU members like France are examining if Grok’s images breach local laws.

Late Wednesday, xAI announced changes to stop Grok users from altering images of real people wearing revealing clothes like bikinis.

xAI stated it restricts users by location from creating images of people in revealing clothing where it is illegal, but did not specify which areas.

Since then, it has limited its image-generation features and placed them behind a paywall.

On Wednesday, Elon Musk stated he did not know of any nude images of minors created by Grok. 

He said he’s “not aware of any naked underage images generated by Grok. Literally zero. Obviously, Grok does not spontaneously generate images, it does so only according to user requests.”

About one image is uploaded every minute on X, according to Copyleaks. Genevieve Oh, an independent internet researcher, reviewed @Grok’s posts on X for 24 hours. About 6,700 images per hour generated by the chatbot were flagged as sexually suggestive or nudifying.