Google Launches AI Detector Portal to Identify Deepfakes Using SynthID

Google has launched a new online tool, the SynthID Detector Portal, aimed at identifying AI-generated media through invisible watermarks embedded in content created with its own tools.

Credit: GOOGLE

The portal allows users to upload files and scan them for SynthID, a watermarking technology developed by Google. It highlights areas within images that have likely been modified by AI, helping users pinpoint synthetic changes in otherwise authentic media.

SynthID watermarks are automatically embedded in media generated by Google’s AI tools, including Gemini for text and images, Veo for video, and Lyria for audio. These watermarks are designed to survive basic edits and remain detectable even after being shared online or through messaging platforms.

The portal builds on recent updates to Google Photos, where SynthID detection was added to identify edits made using the Magic Editor feature.

While the SynthID Detector Portal is a step forward in combating AI-driven misinformation, it has limitations. It can only detect content that includes Google’s SynthID watermark and cannot identify media generated by other platforms, such as ChatGPT, that do not use the technology.

To expand SynthID’s reach, Google is forming partnerships with third parties. In March, it collaborated with Nvidia, and on Tuesday, it announced a new partnership with content verification company GetReal, which will integrate SynthID detection into its existing verification tools.

The SynthID Detector Portal is available to early testers starting Tuesday, May 20. Journalists, media professionals and researchers can join a waitlist to request access.

---

• Google launches SynthID Detector Portal to identify AI-generated media

• Tool scans for invisible watermarks in content created with Google’s AI tools

• SynthID survives basic edits and online sharing

Source: FORBES