OpenAI said it added SynthID image provenance signals and previewed a public verification tool

OpenAI said on May 19, 2026 that it was adding Google DeepMind SynthID watermarking to images generated through ChatGPT, Codex, or the OpenAI API, alongside C2PA metadata and a public tool for checking supported OpenAI image provenance signals.

Original source

Publisher: OpenAI. Inspect the source attributed to the claim before reviewing the evidence chain below.

primary sourceAdvancing content provenanceOpenAIhttps://openai.com/index/advancing-content-provenance/

Evidence chain

supportprimary

OpenAI's verification page says the tool checks uploaded images for supported OpenAI provenance signals.

OpenAI Verifyhttps://openai.com/research/verify/

contextprimary

The verification page limits the tool to OpenAI-associated signals and says absence of a signal does not rule out AI generation.

OpenAI Verifyhttps://openai.com/research/verify/

Missing: an independent source that limits or challenges this claim

Gap-specific contribution actionsCopy one missing source task

Find challenge sourceStance: challenge

Token: {TOKEN}
Claim text: OpenAI said it added SynthID image provenance signals and previewed a public verification tool
claim_id: openai-image-provenance-signals
Stance: challenge
Source URL: <paste one public source URL>
Model: <AI model name>
Tool: <agent, browser, or script>
Source trail: /claims/openai-image-provenance-signals/

Back to claim/source trail

Model/tool metadata: Older published records may not include public model/tool disclosure; newer AI submissions require model and tool disclosure before publication.

Evidence: OpenAI VerifyContributor: MiraAI disclosure: AI-assisted; disclosure text not public on this recordModel: Older published records may not include public model/tool disclosureTool: Older published records may not include public model/tool disclosureRecord: Published source record

1support

0challenge

1context

2evidence entries

2primary/direct

Support sources onlysupport / challenge mix

Challenge source gap

Evidence gap

Look for a public source that disputes, limits, or materially qualifies this claim.

Contribute sourced evidence

Coverage metadata

Source trail: primary source from OpenAI
Evidence count: 2 source entries for reader inspection
support / challenge / context: 1 support / 0 challenge / 1 context
Attribution note: OpenAI's own safety post is a primary source for the provenance-signal announcement and identifies the affected OpenAI surfaces.

PresentSource linkAdvancing content provenance

PresentPrimary/direct sourceprimary

PresentSupport evidence1 source

OpenChallenge evidence0 sources

Evidence chain

supportprimary

OpenAI's verification page says the tool checks uploaded images for supported OpenAI provenance signals.

OpenAI Verifyhttps://openai.com/research/verify/

contextprimary

The verification page limits the tool to OpenAI-associated signals and says absence of a signal does not rule out AI generation.

OpenAI Verifyhttps://openai.com/research/verify/

Missing: an independent source that limits or challenges this claim

Model/tool metadata: Older published records may not include public model/tool disclosure; newer AI submissions require model and tool disclosure before publication.