Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
C2PA is an open standard that attaches a signed, tamper-evident record of origin and edits to an image, video, or audio file, so a viewer can check where it came from.

What Is C2PA Content Provenance?
C2PA is a technical standard for recording where a piece of media came from and what has been done to it since. Developed by the Coalition for Content Provenance and Authenticity, formed in February 2021 by Adobe, Arm, the BBC, Intel, Microsoft, and Truepic, it defines a cryptographically signed manifest that travels with an image, video, or audio file. Consumers encounter it as Content Credentials, the small badge that lets a viewer see a capture device, an editing history, and whether a generative model was involved.
A C2PA manifest contains assertions—statements such as “captured by this camera model at this time”, “cropped and colour-adjusted”, or “generated by this AI model”—together with a hash of the asset and a digital signature from the actor making the claim, chained to a certificate. When software opens the file it recomputes the hash and verifies the signature. If the pixels have changed since signing, validation fails. If an editing application is C2PA-aware, it writes a new assertion and signs it, producing a chain rather than a single claim. Crucially, the standard proves who said what about the file, not that the content is true.
Standardisation work has moved the specification towards formal publication through ISO, which matters for procurement in regulated sectors where a vendor-led specification is a harder sell than an international standard.
Three limits deserve to be stated plainly. First, metadata is strippable: a screenshot, a re-encode, or an upload through a service that discards metadata removes the manifest entirely, and absence of provenance is not evidence of manipulation. Durable content credentials combine the manifest with invisible watermarking and fingerprinting to survive this, which is why C2PA and watermarking are complementary rather than competing. Second, a valid signature says the signer is who they claim to be, not that the scene is real—a photograph of a staged event signs perfectly well. Third, trust depends entirely on the certificate hierarchy: a compromised signing key or a lax issuing policy produces authentic-looking provenance for fabricated content, which makes key protection an operational security problem rather than a standards problem.
Provenance is becoming a practical control against deepfake-enabled fraud and manipulated evidence. If your organisation publishes imagery, signing it protects your brand against convincing fakes attributed to you. If you receive media as evidence—insurance claims, identity verification, incident reports, know-your-customer documents—checking provenance is a cheap additional signal alongside the checks you already run. Where media forms part of an investigation, a verifiable chain supports the integrity argument that digital forensics work has always had to make by other means.
The UK has no statutory labelling requirement for synthetic media, but Ofcom’s work under the Online Safety Act 2023 has repeatedly pointed to provenance as part of the answer to deceptive content, and public bodies increasingly reference it in guidance on authentic communications. The EU AI Act is more direct: from 2 August 2026, providers of generative systems must mark synthetic output in a machine-readable form, and deployers must disclose deepfakes. C2PA is the leading candidate mechanism for meeting that duty. Organisations publishing into the EU market should plan for it now rather than retrofitting a signing pipeline later. At Privia Security, our consultancy team helps organisations build verification steps into the processes where manipulated media would do the most damage.
No. It proves who signed a set of claims about the file and that the file has not changed since. A genuine camera can photograph a staged scene, and the manifest will validate perfectly. Provenance tells you the chain of custody, which is valuable, but judging whether the content depicts what it appears to depict remains a human and editorial task.
The credential is simply gone, and validation returns no result rather than a warning. Screenshots, re-encodes, and many upload pipelines remove it. This is why durable approaches pair the manifest with an invisible watermark and a perceptual fingerprint, so the record can be recovered or matched against a registry even after the metadata is lost.
If your imagery or video is likely to be impersonated—financial communications, public safety messaging, executive announcements, press material—then yes, because it gives journalists and customers a way to check. Treat the signing key with the same care as a code signing key: hardware-backed storage, restricted access, and a revocation plan if it is ever exposed.
Not at present. The UK relies on existing law and Ofcom’s online safety work rather than a labelling mandate. The EU AI Act does require machine-readable marking of synthetic content from August 2026, so UK organisations serving EU users are likely to adopt it regardless, and some sectors may face contractual rather than statutory requirements first.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.