{"data":{"id":"aa63b94d-9447-4ec7-93a3-029d6d0019a7","title":"Advancing content provenance for a safer, more transparent AI ecosystem","summary":"OpenAI is improving how people can verify where AI-generated images and audio come from by using multiple approaches: adding C2PA conformance (a cross-industry standard using metadata and cryptographic signatures to attach source information to content), partnering with Google to embed invisible watermarks called SynthID into images, and releasing a public tool to verify if images came from OpenAI. These layered approaches help protect provenance information (details about content's origin and creation) even when it's edited, downloaded, or shared across different platforms.","solution":"The source describes OpenAI's implemented approaches rather than fixes to a problem. OpenAI has: (1) become C2PA Conforming, which gives platforms a 'trusted way to read, preserve, and pass along the provenance information' attached to content; (2) incorporated 'SynthID embeds an invisible watermarking layer that complements C2PA metadata-based approaches,' starting with images from ChatGPT, Codex, or the OpenAI API; and (3) is 'previewing a' public verification tool for users to detect if images came from OpenAI. The source states these approaches are designed to work together: 'C2PA helps content carry detailed context; SynthID helps preserve a signal when metadata does not survive.'","labels":["safety","policy"],"sourceUrl":"https://openai.com/index/advancing-content-provenance","publishedAt":"2026-05-19T10:45:00.000Z","cveId":null,"cweIds":null,"cvssScore":null,"cvssSeverity":null,"severity":"info","attackType":[],"issueType":"news","affectedPackages":null,"affectedVendors":["OpenAI","Google"],"affectedVendorsRaw":["OpenAI","Google DeepMind","DALL-E 3","Sora","ChatGPT","Codex","Voice Engine"],"classifierModel":"claude-haiku-4-5-20251001","classifierPromptVersion":"v3","cvssVector":null,"attackVector":null,"attackComplexity":null,"privilegesRequired":null,"userInteraction":null,"exploitMaturity":null,"epssScore":null,"patchAvailable":null,"disclosureDate":"2026-05-19T10:45:00.000Z","capecIds":null,"crossRefCount":0,"attackSophistication":"moderate","impactType":["integrity"],"aiComponentTargeted":"model","llmSpecific":false,"classifierConfidence":0.85,"researchCategory":null,"atlasIds":null}}