OpenAI has announced two complementary measures to help users identify AI-generated images, partnering with Google on technology that addresses one of the most pressing challenges in the era of generative AI.
The company is adopting C2PA, an open standard developed by a non-profit coalition founded in 2021, which embeds readable metadata signals identifying an image as AI-generated. Alongside this, OpenAI is integrating Google’s SynthID, an invisible watermark engineered to survive manipulation attempts including screenshots, resizing, and digital editing — making it significantly harder for bad actors to erase.
OpenAI is also previewing a public verification tool that checks for both signals simultaneously. The company said the two systems are deliberately designed to offset each other’s weaknesses, with metadata providing richer information and watermarking offering greater durability. Coverage will initially be limited to OpenAI-generated images, with broader expansion planned.
