**[OpenAI Teams Up with Rival Google: ChatGPT Fully Integrates DeepMind's Invisible Watermark and Launches Image Authenticity Verification Tool]** According to monitoring by Beating, OpenAI has announced a collaboration with former rival Google DeepMind to fully integrate SynthID invisible watermark technology into ChatGPT, Codex, and API-generated images. Additionally, OpenAI has officially become a compliant product under the C2PA image credential standard and has launched a public preview version of its image authenticity verification tool. The C2PA metadata standard relies on cryptographic signatures to record the source of images, but this metadata is prone to loss when users download, screenshot, or convert image formats. To address this vulnerability, OpenAI has adopted SynthID invisible watermark technology developed by Google DeepMind, embedding imperceptible identifiers directly into image pixels. Even if images undergo secondary processing such as screenshots, SynthID can still persist. This dual-layer defense mechanism combines the rich contextual information of metadata with the durability of invisible watermarks. As part of this update, OpenAI has also launched a website for verifying image authenticity ([openai.com/research/verify](openai.com/research/verify)), which is now open for public preview. Users can directly upload images via the webpage, where the tool will detect whether the image contains OpenAI's Content Credentials or SynthID invisible watermark. However, OpenAI has adopted a cautious stance regarding detection results. Since watermarks and metadata can still be forcibly removed under extreme circumstances, the tool cannot definitively conclude that an image is not AI-generated even if no anti-counterfeiting signals are detected. [Original Link]