In a world where it is frighteningly easy to manipulate media content, fighting fake news and validating the source of that content to prevent it from being altered is vital. This paper introduces the key concepts of blockchain and smart contracts and proposes a design to ensure the traceability of media assets in a potentially untrustworthy environment.
In this particular use-case, the distributed ledger keeps track of each operation which happens to an asset from its capture through to publication, including editing. This information can be checked against the ledger, in order to validate intermediate steps required to produce the resulting asset. This design ensures integrity of media assets including reliable chronological dating, and the validation of sources while preserving anonymity. At each stage of the workflow, manual or automated verifications are recorded, which creates chains of trust without central authority.
It is all too familiar: a news report appears online with questionable origins; press reports of social media storms based on falsely attribute reports. It’s becoming important to be able to distinguish genuine content and to identify the sources of that content.
With modern editing tools, almost any individual is capable of editing video content at high quality, and immediately accessing an audience of billions through modern social networks – a potentially dangerous development for the media industry.
When the infrastructure required to capture, edit and produce high quality content was the preserve of large media institutions, modification of that content was the preserve of jokes and novelty pranks.
These trusted organisations also had exclusivity in being able to target large audiences rapidly. The term “viral” was an exclusively medical term 15 years ago.
Many examples of this phenomenon exist today. The BBC was forced to deny a story perhaps produced as a prank, but which went viral as outlined in the Telegraph, Horner.
Manipulated images will become more and more difficult to detect as shown by Suwajanakorn et al. They demonstrated how they used former US President Barack Obama’s videos to synthetise his face and how they learned from audio how to fake lip-sync on a different video with the same audio.
News material has therefore to be traceable. It’s not just the date and location of shooting, but a verified source and content accuracy are also now priorities for any media institution.
This paper will introduce the key concepts of blockchain and propose a design of public blockchain to ensure the traceability of media assets from production to distribution in a potentially untrustworthy environment.