At Google I/O 2026, Sundar Pichai announced that OpenAI, ElevenLabs, and Nvidia are adopting Google’s SynthID invisible watermarking technology. Google also open-sourced SynthID’s text watermarking. The coalition now represents the most significant cross-industry content provenance standard, with SynthID integration expanding to Google Search and Chrome.
What Is SynthID?
SynthID is Google’s invisible watermarking technology that embeds imperceptible signals in AI-generated images, video, and audio. These watermarks survive normal editing and compression and can be verified by detection tools to determine whether content originated from AI or a camera, and whether it has been edited with generative AI tools.
Who Is Adopting SynthID?
The adoption represents a who’s-who of the AI industry. OpenAI’s integration begins immediately with images created by ChatGPT, Codex, and the API — the most significant partnership given OpenAI and Google’s direct competition. ElevenLabs extends provenance to AI audio, and Nvidia covers AI video generated by its Cosmos platform. South Korean internet giant Kakao is also adopting the standard.
How Will SynthID Reach Billions of Users?
SynthID verification is expanding to Google Search and Chrome browser in coming weeks, meaning billions of users will encounter AI content labeling in their everyday digital experience. C2PA Content Credentials — the broader industry provenance standard that SynthID integrates with — is also rolling out in the Gemini app.
Why Is This a Big Deal?
This is the most significant AI governance development of the month. The combination of OpenAI’s adoption (making SynthID present in the world’s most used AI image generator), ElevenLabs (AI audio), and Nvidia (AI video) creates a content provenance infrastructure of unprecedented breadth. Google also open-sourced SynthID’s text watermarking technology, making it available to any developer.
Key Takeaways
- OpenAI, ElevenLabs, Nvidia, and Kakao adopting Google’s SynthID watermarking
- OpenAI integration starts immediately with ChatGPT, Codex, and API images
- Google open-sourced SynthID text watermarking for all developers
- SynthID verification coming to Google Search and Chrome
- C2PA Content Credentials rolling out in Gemini app
- Creates the broadest cross-industry content provenance standard to date
Frequently Asked Questions
Can SynthID watermarks be removed? SynthID watermarks are designed to survive common editing and compression, but determined actors can attempt to strip them. The standard is a significant deterrent, not an absolute solution.
Does this mean Google controls AI content labeling? While SynthID is Google’s technology, the open-source release of text watermarking and integration with C2PA (an independent standard) means the provenance ecosystem is becoming more open, not more centralized.
How do users check for SynthID watermarks? Google is adding verification tools to Search and Chrome, and OpenAI has launched a public image verifier for images made with its platforms.