Google Photos is introducing a new function that uses artificial intelligence to detect edited images. The Reimagine tool now embeds an invisible watermark, known as SynthID, into photos that have been edited by AI. This technology from Google DeepMind is designed to enhance visibility when an image has been altered.
Summary of essential information to understand.
- Google Photos is now incorporating a hidden watermark into images that have been edited using AI, utilizing SynthID technology.
- The digital watermark is invisible to the naked eye but can be identified using certain tools.
- Reimagine is exclusively accessible on Pixel 9 or newer devices; however, the images with the watermark can be viewed by anyone.
Google uses artificial intelligence to insert undetectable watermarks into photos.
Users are now able to have greater influence over distinguishing genuine images from those altered by AI through Reimagine. Google has introduced a new feature that includes a notice stating “Credit: Edited with Google AI” in the image details when an image is modified, aiding in the identification of AI-induced alterations.
Users can now identify images generated or altered by AI algorithms with a new feature. It is important to note that while the invisible watermark system is hard to erase, it can be compromised by minor alterations or multiple edits, reducing its reliability in some situations.
External tools are currently unable to identify watermarks, making it difficult for platforms like Instagram to detect AI-generated or edited images without access to this technology from Google.
Source