March 11, 2026, 8:25 am | Read time: 2 minutes
AI-altered images are increasingly appearing on X. The platform is now responding with a new setting for photo uploads. It aims to prevent manipulations and potential sexualization by the chatbot Grok.
The microblogging service X is trying to better protect users from AI manipulations of their photos. This comes in response to cases where images were altered using the chatbot Grok. Reports indicate that this feature was misused in early January to fake and sexualize photos of men, women, and children. The platform is now reacting with a new setting that can be activated directly when uploading an image. It is designed to prevent Grok from automatically editing a photo.
New Setting for Image Uploads
When uploading an image, you can activate a switch that is intended to block modifications by Grok. The feature is labeled “block modifications by Grok.” As reported by “Social Media Today,” the setting is currently available in the iOS app. It is located directly in the options during image upload. Users can decide whether their image can be altered by the AI chatbot or not. The new setting is part of X’s measures to address criticism surrounding AI manipulations.
EU Intensifies Fight Against Cyberbullying With an App
How to Check if Your Photo Is Being Used for Deepfakes
Grok Update with Weaknesses
However, the protection is limited. The option only prevents the @Grok account from being immediately tagged under an image to initiate editing. Following international outcry, X had already restricted this possibility. For free X accounts, direct use is now limited. Users with a premium subscription can still tag Grok and have photos edited. Thus, the feature remains fundamentally available, even though the new switch is intended to introduce additional restrictions.
Restrictions Can Be Bypassed
Tests show that the new block can be partially bypassed. In the app, users can press and hold an image and then select “Edit image with Grok.” The photo opens directly in the Grok app. There is also another alternative: A protected image can be downloaded and then re-uploaded. It can then be edited by Grok again. Additionally, the new setting does not yet appear during uploads via the web browser or for images previously published on the platform.
The option is currently visible only in the iOS app and was apparently introduced without a major announcement. It is not yet known if or when an Android version will be released.