OpenAI, Google DeepMind Launch New AI Verification System

United States, May 20, 2026 - OpenAI, the company behind ChatGPT, has unveiled a series of new verification technologies and industry partnerships aimed at curbing misinformation and manipulated media generated through AI chatbots.

 In a statement, the company has said that it has partnered with various companies, including Google, to improve Provenance signals, which give people more context about where the content came from, how it was created or edited, and whether it is what it claims to be.

The move comes as millions of people globally are leveraging AI tools to create and edit images, audio, and other forms of media for communication and online sharing.

 While some outlets and individuals disclose when they use AI to generate content, others do not, making people more vulnerable to false information online.

OpenAI has said that it has now become a “C2PA Conforming Generator Product,” which means that platforms and publishers will be able to more easily recognize, preserve, and pass along provenance information attached to AI-generated media.

OpenAI noted that the provenance data leverages metadata and cryptographic signatures to securely carry information about how media was created or edited, thus helping users better assess digital content.

Through the partnership with Google DeepMind, the company has revealed that it has added more durable SynthID watermarking to images that have been produced by ChatGPT, Codex, and the  OpenAI API.

OpenAI has indicated that, unlike traditional metadata, which can be erased during uploads, downloads, screenshots, or file conversions, the SynthID is embedded with an invisible watermarking signal directly to images, which will make procenance tracking more efficient.

“These two systems reinforce each other. C2PA helps content carry detailed context; SynthID helps preserve a signal when metadata does not survive. Watermarking can be more durable through transformations like screenshots, while metadata can provide more information than a watermark alone,” the company stated.

 The company is currently in the process of previewing a new public verification tool, which will enable people to upload and verify whether an image was generated by a chatbot by checking whether it contains provenance signals, including Content Credentials and SynthID.

“At launch, the tool is limited to content generated by OpenAI. In the upcoming months, we aim to support cross-industry efforts to make verification possible across platforms,” the company stated.