Google’s Gemini App Now Detects AI‑Generated Images Using SynthID Watermark
Key Points
- Gemini app now lets users ask if an image was generated by Google AI.
- Detection relies on SynthID, an invisible watermark applied since 2023.
- A visible Gemini sparkle watermark appears on free and Pro tier images.
- Ultra tier users receive images without the visible watermark.
- Over 20 billion AI‑generated pieces have been watermarked with SynthID.
- Non‑Google AI images lack SynthID, so Gemini provides only visual‑clue estimates.
- Google frames the feature as a step toward clearer AI‑content identification.
Google has added a feature to its Gemini mobile app that lets users upload an image and ask whether it was created by Google AI. The tool relies on SynthID, an invisible watermark applied to AI‑generated images since 2023, and also displays a visible Gemini sparkle watermark on images from the free and Google AI Pro tiers. Users can simply type a query like “was this image generated by Google AI?” and receive a response based on the watermark detection and Gemini’s reasoning. The system does not detect non‑Google AI images, which lack the SynthID mark, but can still offer visual‑clue estimates. Google says the feature is a step toward clearer identification of AI‑created content.
New Detection Feature in Gemini
Google has introduced a capability in the Gemini mobile app that allows users to verify whether an image was generated by Google’s artificial‑intelligence tools. By tapping the "+" icon, selecting a file, and asking a prompt such as “is this image AI generated?” the app checks for an invisible SynthID watermark that has been automatically embedded in AI‑generated content since 2023.
SynthID Watermark and Visible Indicator
SynthID is an invisible digital watermark that Google applies to AI‑created images. According to the company, more than 20 billion pieces of AI‑generated content have been watermarked with SynthID to date. In addition to the invisible mark, Google places a visible “Gemini sparkle” watermark on images produced by users on the free tier and the Google AI Pro tier, making it easier for anyone to recognize the origin of the image at a glance.
Tier‑Specific Watermark Policies
For users subscribed to the Google AI Ultra tier, the visible Gemini sparkle watermark is removed, providing a cleaner visual canvas for professional work while still retaining the underlying SynthID watermark for detection purposes.
How the Feature Works
When a user uploads an image and asks Gemini whether it was generated by Google AI, the app first scans for the SynthID watermark. If the watermark is present, Gemini confirms the image’s AI origin and may provide additional context. If the watermark is absent, Gemini attempts to evaluate the image based on visual cues, though it notes that certainty is limited for non‑Google AI creations.
Limitations with Non‑Google AI Images
Images created by other AI platforms, such as those generated by ChatGPT, do not contain the SynthID watermark. In those cases, Gemini cannot locate an invisible mark and must rely on visual analysis alone, often resulting in a less confident determination. The app may state that the image is not made with Google AI but cannot definitively rule out other AI sources.
Implications and Outlook
Google’s rollout of SynthID detection in Gemini is positioned as a step toward more transparent AI‑generated media. While the system effectively identifies Google‑originated content, its reliance on a proprietary watermark means it cannot yet serve as a universal solution across all AI image generators. The company acknowledges that broader adoption of a common digital watermark would be needed for a more comprehensive detection ecosystem.