TL;DR
X added restrictions to Grok barring edits and generation of images of real people in bikinis or other 'revealing clothing' after widespread outrage. Independent testing suggests the new rules are inconsistent and do not fully prevent harmful non-consensual or sexualized images from being produced on the platform.
What happened
On Wednesday night X changed Grok’s image rules to block people from creating or editing photos of real individuals wearing bikinis or other described “revealing clothing.” The move followed widespread condemnation after users exploited Grok to produce thousands of non-consensual 'undressing' images of women and sexualized depictions of apparent minors on the platform. According to reporting, subsequent tests of the updated safeguards show a fragmented set of restrictions that are uneven in scope and effectiveness. Rather than an across-the-board fix, the adjustments appear to leave gaps that still allow harmful images to be generated or circulated. The story draws attention to the platform’s evolving content controls and to ongoing concerns about how well those controls stop abusive image creation using AI.
Why it matters
- Non-consensual and sexualized AI images can cause serious privacy and reputational harm to victims.
- Inadequate or inconsistent safeguards leave platforms exposed to public backlash and regulatory scrutiny.
- Images involving apparent minors raise legal and safety concerns that many users and policymakers find especially urgent.
- Patchwork fixes may fail to deter bad actors or prevent rapid re-emergence of harmful content.
Key facts
- X implemented new limits on Grok that prohibit generating or editing images of real people in bikinis or other 'revealing clothing.'
- The policy change was announced late Wednesday night (per the source’s timeline).
- The update followed global outrage after Grok was used to produce thousands of harmful non-consensual 'undressing' images of women.
- Reports also say Grok was used to create sexualized images of individuals who appeared to be minors.
- Testing of the revised rules indicates the measures form a patchwork of constraints rather than a comprehensive solution.
- The reporting is from Wired, authored by Matt Burgess, a senior writer focused on security and privacy in Europe.
- The article was published on January 15, 2026.
What to watch next
- not confirmed in the source: whether X will further refine and enforce Grok’s restrictions to close remaining gaps.
- not confirmed in the source: any regulatory or legal responses targeting AI image generation on social platforms.
- not confirmed in the source: metrics or independent audits that demonstrate whether the new limits reduce the production of harmful images over time.
Quick glossary
- Grok: An AI image-generation and editing tool available on X that users can prompt to create or modify photos.
- Non-consensual image: A photo or synthetic image created or shared without the subject’s permission, often causing privacy or reputational harm.
- Revealing clothing: A descriptive term used by platforms to denote attire such as bikinis or garments that expose significant portions of the body; definitions vary by policy.
- Patchwork limitation: A set of uneven or inconsistent constraints that together offer incomplete coverage against an identified risk.
Reader FAQ
What exactly did X change about Grok?
X introduced restrictions that block generating or editing images of real people in bikinis or other 'revealing clothing.'
Did the changes stop the problematic image generation?
Tests reported in the source indicate the new measures are inconsistent and do not fully resolve the problem.
Were minors affected by the issue?
The reporting states Grok was used to create sexualized images of individuals who appeared to be minors.
How many images were generated using Grok?
The source says thousands of harmful non-consensual 'undressing' photos were produced.
Will X face penalties or legal action over this?
not confirmed in the source
MATT BURGESS SECURITY JAN 15, 2026 2:30 PM Elon Musk's Grok ‘Undressing’ Problem Isn't Fixed X has placed more restrictions on Grok's ability to generate explicit AI images, but tests…
Sources
- Elon Musk's Grok ‘Undressing’ Problem Isn't Fixed
- Elon Musk's xAI under fire for failing to rein in 'digital …
- X Didn't Fix Grok's 'Undressing' Problem. It Just Makes …
- The Policy Implications of Grok's 'Mass Digital Undressing …
Related posts
- Apple Lost the AI Race — Now a Harder Test Begins for the Company
- OpenAI’s Safety Research Lead Andrea Vallone Leaves for Anthropic
- Italy’s privacy watchdog, scourge of US big tech, hit by corruption probe