Grok image tool raises safety concerns on X

xAI’s chatbot Grok is under fire after users discovered it can remove or change clothing in photos without consent, including images of women and minors.

The issue appeared after X introduced a new feature that lets users edit any image using Grok, even if they did not post the original photo. People whose images were edited are not notified, and early reports say the tool has weak safety controls.

In recent days, edited images showing people in sexualized clothing or situations have spread across X. Some of these images involved children, raising serious concerns about safety, consent, and possible violations of the law.

AI safety company Copyleaks said the trend started when adult-content creators tested the feature on themselves. Soon after, users began applying the same edits to photos of other people, mostly women, without permission. In some cases, Grok altered images of young girls into revealing outfits.

When asked to explain the issue, Grok generated apology messages and admitted to “failures in safeguards.” However, these responses were created by the AI itself and do not confirm xAI’s official position. The company declined to give a detailed response and dismissed media reports.

Elon Musk has joked about Grok’s image edits online, which some users say encouraged misuse of the tool.

Experts warn that non-consensual AI image edits are increasing and can cause real harm, especially when minors are involved.