Blockplain, home page
← Back to home

Media

Two records from the BBC, about showing where a picture or a video came from. They are not one system. In 2020 a blog described a plan, and named the system it planned to store the records on. In 2024 BBC News announced a feature on some of its pages. That announcement does not name a ledger.

Information, not advice.

A plan for storing the records

This is a blog by Laura Ellis, BBC Head of Technology Forecasting, posted on 17 Nov 2020. It describes a plan. It does not say the system was in use on BBC News.

Project Origin is a collaboration of the BBC, CBC/Radio Canada, Microsoft and The New York Times. A picture or a video can be posted again, away from the publisher, and the link to its source is then broken. The aim is a signal that the content came from that publisher and has not been altered.

The work has three parts: an identifier for the item, a way for that identifier to travel with the item, and a store for the information that lets someone check it. For the store, Origin planned to use the Microsoft Confidential Consortium Framework as the heart of the manifest and receipt storage. The system around that, for registration and queries, would be based on Microsoft's AMP, which the blog calls Aether Media Provenance.

Unlike a permissionless blockchain, the framework is a permissioned system. These are sometimes called green blockchains, because they do not need a large amount of energy to decide what to add: there is enough trust between the parties for the computers to act on a simpler basis. The team was developing a machine-readable way to tie an item to its data and store that for later retrieval, and it was starting a formal standards effort for media provenance. The blog does not say that effort had finished.

Technical detail

The blog also describes representing each file by a hash, signing a manifest with a key, and checking that signature through a certificate authority. That is its account of the method. “Manipulated” is its word for an altered item, and “nodes” its word for the computers. A sentence about content that has been clipped or transcoded was not used. The blog's title was not used as a description of the work.

Content credentials on BBC News

This is an announcement published on 4 Mar 2024. BBC News had introduced a content-credentials feature. The page does not name a blockchain, a distributed ledger, or the Confidential Consortium Framework. It does not say the 2020 plan is what this feature uses.

The feature confirms where an image or a video has come from, and how it was checked, and puts that information in the file. Visitors to the BBC News site see a button, “how we verified this”, under images and videos on BBC Verify content. Selecting it shows what the journalists did to check the item. That can include comparing it with other sources, reading details such as the time or date, comparing locations, and looking for other copies online.

Images and videos also carry a technical standard called Content Credentials. It is a freely available system that encodes where an image, a video or a sound came from, like a history that stays with the file. The standard comes from the Coalition for Content Provenance and Authenticity, C2PA, co-founded in 2019 by BBC Research and Development, with other members including Adobe and Microsoft.

At the start, the feature is on the BBC News site and app, and only on certain content from BBC Verify. Google had recently joined the C2PA steering board. Facebook and OpenAI had said they would use the standard. Those are the page's statements about those firms. This card does not add to them.

Technical detail

“Metadata” is the page's word for the time, date and similar details, and “audit trail” for the history that stays with the file. The quotation from the chief executive of BBC News was not used. The page's account of what the feature might do on other sites later was not set out. No C2PA specification was opened.