OpenAI and Google Expand Image Provenance Checks with C2PA and SynthID

Overview

OpenAI and Google both announced image provenance updates on May 19, 2026. The shared direction is clear: major image platforms are moving beyond simple labels toward a layered system that combines signed metadata, watermarking, and consumer-facing verification tools.

OpenAI says it is strengthening provenance for AI-generated media through C2PA conformance, SynthID image watermarking, and an early public verification tool. Google says it is expanding content transparency features across Search, Gemini, Chrome, Pixel, and Cloud, including verification for SynthID and C2PA Content Credentials.

Content Credentials

C2PA Content Credentials are signed provenance records that can travel with a digital asset. OpenAI says becoming a C2PA Conforming Generator Product gives platforms a more trusted way to read, preserve, and pass along provenance information attached to OpenAI-generated content.

Google says it already uses C2PA Content Credentials across a growing number of generative media tools. Google also points to Pixel camera capture as part of the same trust chain, saying Pixel 10 was the first smartphone to provide Content Credentials for images in its native camera app and that video support is expanding to Pixel 8, 9, and 10 phones.

The C2PA explainer is careful about the limits. Content Credentials can make provenance tamper-evident and verifiable against trust information, but they do not decide whether a photo or video is true. That distinction matters for photographers, editors, and anyone handling images from unknown sources.

SynthID Watermarking

OpenAI says C2PA metadata is important, but can be stripped or broken when media is uploaded, downloaded, converted, resized, or captured through screenshots. To make provenance more durable, OpenAI is adding Google DeepMind SynthID watermarking to images generated through its tools.

Google describes SynthID as an imperceptible watermarking technology already used across its generative media models and products. In this workflow, watermarking does not replace metadata; it provides another signal that may survive cases where ordinary metadata is lost.

Verification Tools

OpenAI is previewing a public tool that checks uploaded images for provenance signals including Content Credentials and SynthID. OpenAI says the tool is initially limited to OpenAI-generated content and should not make a definitive claim when no signal is detected because provenance data can be missing or stripped.

Google says SynthID verification for image, video, and audio is being expanded from Gemini to Search and Chrome. Google also says C2PA Content Credentials verification is rolling out in Gemini, with Search and Chrome support planned in the coming months.

Why It Matters

PhotoTools readers often work at the exact points where provenance can change: converting formats, resizing, compressing, cropping, removing EXIF, and preparing images for upload. Those actions can be useful for privacy and file size, but they can also affect metadata-based provenance.

That does not mean every image should keep every metadata field. A privacy-first workflow may deliberately remove EXIF or GPS data, while a newsroom, marketplace, or contest workflow may want to preserve selected Content Credentials. The practical takeaway is that image tools increasingly need to treat metadata as a user-facing trust and privacy decision, not just hidden file baggage.