A recent report from Android Authority reveals that OpenAI may be planning to introduce watermarks on images generated by ChatGPT in a future update. The discovery stems from a teardown of the app’s latest beta release (v1.2025.196), where a new option labeled “Save without watermark” was spotted in the app’s settings menu. This option, located behind the three-dot icon in the top-right corner of image generation results, suggests that watermarks could soon become the new default for saved images — with the ability to disable them under certain conditions.
This change, if implemented, would represent a significant step toward increased transparency in the distribution of AI-generated content. With growing concerns about deepfakes, misinformation, and unauthorized use of generative tools, watermarking could help make it easier to trace the origin of synthetic media — especially in contexts where authenticity is crucial.
However, it’s still unclear whether all users will have access to the “Save without watermark” option. Android Authority speculates that this feature might be restricted to paid users — likely as part of ChatGPT’s Plus subscription tier — meaning free users would only be able to download images that visibly indicate they were generated by AI. This kind of tiered control could act as both a monetization strategy and a safeguard to encourage responsible image sharing.
At this stage, OpenAI has not made any formal announcement regarding the feature, and there’s no guarantee that watermarking will appear in the final release version of the app. Still, its presence in the beta code points toward a possible future where AI-generated content carries a visual signature by default — a move that could influence how generative imagery is used across social platforms, creative projects, and professional settings.