Microsoft Paint, Photos Hide AI Watermarks, Report Finds
Microsoft's Paint and Photos apps quietly stamp AI-made images with a hidden tracking code, a developer has revealed. If you create or share AI visuals through these tools, that image may already carry an invisible identifier you never agreed to.
What actually happened
Tom's Hardware reports that an independent developer decoded a hidden watermarking system built into Microsoft Paint and Photos. The system embeds a GUID, a unique code string, inside images made with the apps' built-in AI tools. The outlet does not name the developer or confirm which Windows or app versions carry the feature. It is unclear if every AI-generated image gets tagged, or only some. Microsoft has not commented publicly, based on what has been reported so far. The exact contents of the GUID and how anyone could read it back out remain unknown. That silence leaves open questions about scope and intent.
How we got here
Watermarking AI content has become routine as image generators spread through everyday software. Microsoft folded AI generation directly into Paint and Photos, giving it an obvious spot to slip in provenance markers. Nothing in current reporting says when this began or whether Microsoft ever disclosed it. There is also no comparison yet to other watermarking standards used across the industry. What stands out is simple: this tagging ran silently until one developer stumbled onto it. That gap between quiet deployment and public disclosure is becoming a pattern across mainstream AI tools.
Why this matters for you
For everyday users, any image made through Paint or Photos AI features may already hold a hidden marker, seen or not. For people managing digital identity or reputation through wallets and on-chain profiles, this is a reminder that AI-generated content can carry origin data outside their control. Builders working on content verification, including tools tied to bonuz's wearable and AR roadmap, now have another real-world case of embedded tracking to study. Sharing these images, even after cropping or converting them, may not remove the hidden tag.
The bigger question
If a watermark can identify AI-made images without the creator's knowledge, who should hold power over that hidden data? The platform that built it, the person who made the image, or an outside standard nobody controls alone? As AI visuals spread across messaging apps, social platforms, and wearable devices, that question grows harder to sidestep.
What to watch
Microsoft has not confirmed or explained the watermark publicly. Watch for any statement, documentation update, or expansion of this feature to other apps. Further reverse engineering from independent developers could reveal more about how the GUID is structured. As AI-generated visuals move onto smart glasses and other wearables, silent tagging like this may become standard rather than an exception.
