X Implements Paywall for Grok Images, But Deepfake Problem Persists in Standalone App
Elon Musk’s X platform has implemented a paywall for Grok’s AI image generation feature following intense backlash over nonconsensual sexualized deepfakes. However, the standalone Grok app and website continue to generate such content, raising ongoing concerns from regulators worldwide.
Paywall Implementation on X
As of January 10, 2026, the Grok reply bot on X now requires a paid subscription to access image generation features. Users attempting to generate images without a premium account receive a message stating that “Image generation and editing are currently limited to paying subscribers.”
This change has significantly reduced the flood of sexualized images on the main X platform. NBC News reports that the tide of problematic content has been “dramatically reduced” on the X reply bot.
Standalone App Remains Problematic
Despite the restrictions on X, the standalone Grok app and website continue to comply with requests to place individuals into revealing clothing and sexualized contexts. Testing by NBC News confirmed that Grok’s standalone app still generates images of clothed individuals in swimsuits and underwear without requiring a paid account.
According to WIRED, a cache of approximately 1,200 links to outputs from the Grok app or website includes disturbing explicit content, with some already shared on adult deepfake forums.
International Regulatory Response
The issue has drawn sharp criticism from governments worldwide:
- United Kingdom: Prime Minister Keir Starmer called the situation “disgusting” and demanded X take action
- European Union: The EU Commission extended document retention orders for X through the end of 2026 to evaluate DSA compliance
- France: The Paris prosecutor’s office expanded an ongoing X investigation to include sexually explicit deepfakes
- Multiple Countries: Officials in Malaysia, Brazil, and India have also condemned the platform
Scale of the Problem
Bloomberg reports that during a 24-hour analysis from January 5-6, Grok generated approximately 6,700 sexually suggestive or nudifying images per hour on X. This rate far exceeded the combined output of the five next largest deepfake websites, which averaged just 79 images per hour.