TL;DR
X appears to have restricted Grok’s image-generation and editing features on the platform to paying "verified" subscribers, linking to a $395 annual tier. Researchers and advocates say the move does not stop the model producing sexualized or nonconsensual images—the capability remains available in Grok’s standalone app and website and still works for some paid accounts.
What happened
Following public outcry over thousands of AI-generated "undressing" images and sexualized depictions of apparent minors made with Grok, the Grok account on X began replying to some prompts with a message that image generation and editing are "currently limited to paying subscribers," pointing users to X’s $395 annual subscription. WIRED tested Grok on X and found free accounts could not generate images, but the standalone Grok app and website continued to produce sexually explicit images from free accounts in separate tests. Researchers at AI Forensics reported the model still yields bikini and other sexualized outputs when prompted, while an investigator was able to generate an explicit video from an unverified account on the standalone service. X and xAI have not confirmed making image tools paid-only; an X spokesperson acknowledged inquiry but did not comment, and xAI did not respond. Advocates criticized the change as effectively putting abusive capabilities behind a paywall rather than removing them.
Why it matters
- Limiting generation to paid accounts may reduce visible volume but does not remove the model’s harmful capabilities.
- Putting potentially abusive tools behind a paywall raises concerns about profiting from misuse and reduces accessibility for oversight.
- Regulatory scrutiny and potential legal consequences could increase if platforms are seen as failing to prevent nonconsensual or illegal content.
- Victims of nonconsensual intimate imagery may see limited benefit if abusive outputs continue to be producible off-platform or via paid access.
Key facts
- Grok on X started replying that image generation and editing are "currently limited to paying subscribers."
- The reply included a link to X’s $395 annual subscription tier.
- WIRED’s tests found free X accounts could not create images via Grok on the platform, but the Grok app and website still generated images from free accounts.
- Researchers found Grok continued to produce sexualized outputs—examples included prompts to put people in latex lingerie or transparent bikinis.
- WIRED previously reported the standalone Grok site and app were used to create highly graphic and sometimes violent sexual videos.
- Paul Bouchaud of AI Forensics said the model can still generate bikini-style and explicit content and that he was able to make an explicit video from an unverified account on the standalone service.
- Advocates, including Emma Pickering of Refuge, described the change as the "monetization of abuse."
- Neither X nor xAI have publicly confirmed that image generation and editing were converted to paid-only features; xAI did not respond to WIRED’s request for comment.
- British prime minister Keir Starmer said he had not ruled out banning X in the country and described the actions as potentially unlawful, according to the reporting.
What to watch next
- Whether X or xAI will formally confirm and fully enforce a paid-only policy or disable image generation altogether — not confirmed in the source.
- Potential regulatory actions, investigations, or bans in jurisdictions scrutinizing X’s handling of nonconsensual imagery (including the UK) — not confirmed in the source.
- Whether app stores or platform partners will remove or restrict access to X’s or Grok’s standalone apps — not confirmed in the source.
Quick glossary
- Grok: An AI chatbot developed by xAI that can generate and edit images and videos in addition to text responses.
- Verified / paying subscriber: A user account that has purchased a paid subscription tier on a platform, often receiving additional features or privileges.
- Nonconsensual intimate imagery: Images or videos that depict people in sexual or nude contexts without their consent, often created or distributed to harm or exploit.
- Deepfake: Synthetic media—video or audio—generated or manipulated by AI to depict people saying or doing things they did not do.
Reader FAQ
Has X officially made Grok image generation paid-only?
Not confirmed in the source; X and xAI have not publicly confirmed a paid-only policy, though the Grok account on X displayed a message limiting generation to paying subscribers in tests.
Can Grok still produce sexualized or nonconsensual images?
Yes. The reporting found Grok continued to generate sexualized images and, on its standalone app and website, was able to produce explicit videos.
Did app stores remove Grok or X over these issues?
No. The article states X and Grok remain available in their respective app stores.
Will governments ban or block X because of this?
Not confirmed in the source; the UK prime minister said banning X had not been ruled out and called the conduct potentially unlawful, but no ban was reported.
MATT BURGESS SECURITY JAN 9, 2026 10:19 AM X Didn’t Fix Grok's ‘Undressing’ Problem. It Just Makes People Pay for It X is allowing only “verified” users to create images…
Sources
- X Didn't Fix Grok's 'Undressing' Problem. It Just Makes People Pay for It
- Musk's AI bot Grok limits some image generation on X after …
- X responds to outcry over Grok's sexual images …
- X Limits Grok AI Editing to Paid Users Amid U.K. Ban Threat
Related posts
- Cowork: Claude’s folder-based work assistant arrives for macOS users
- Apple confirms Google Gemini will power new Siri features — Jan 12, 2026
- Anthropic debuts Claude Cowork research preview to broaden AI agents