Google is expanding AI-content verification in the Gemini app through SynthID detection and Content Credentials. The development gives users more ways to inspect whether images, video, and audio were created or modified with Google's AI tools, while making provenance signals easier to surface across Google products.
The important confirmed change is centered on verification, not on a universal setting for controlling a visible Gemini watermark. In its official announcement on identifying AI-generated media online, Google detailed SynthID verification for image, video, and audio in the Gemini app, alongside a broader rollout of Content Credentials verification. For businesses using generative media, the update matters because provenance is increasingly part of content governance, review workflows, and brand-risk management.
What Google added to Gemini
Google's May 19, 2026 announcement describes two connected mechanisms for establishing the origin or AI involvement of media. SynthID is Google's invisible watermarking technology, designed to persist in supported AI-generated content and be detected by verification tools. Content Credentials, based on the C2PA standard, provide provenance information that can help people inspect relevant creation and editing history when the data is available.
