Reuters, in collaboration with Canon, a prominent camera manufacturer, has joined the Content Authenticity Initiative (CAI) to promote the credibility of news images. The CAI is focused on showcasing a system that can verify the authenticity of unaltered photographs. Canon has incorporated a unique camera embedded key that adds a digital signature, GPS location, and other metadata to the image. This data is then sent to Reuters’ database, along with registering the creation of the file on blockchain platforms like Filecoin and Storj.
Once Reuters receives the image, various editing processes are recorded within their internal database. However, all confirmation hashes of the editing processes are recorded on the external Hedera blockchain. This ensures that when readers view the image, they can verify its source from Reuters.
However, this system does have its weaknesses. The computation time for generating digital signatures is lengthy, which is a concern when covering real-time events like wars in Ukraine where speed is crucial. Furthermore, the editing of an image can detach its connection to the original, undermining the credibility. To address this issue, software from FotoWare can be used by customers who purchase the images. This ensures that the editing done to the image is certain and does not deviate from the truth.
The verification approach relies on publicly accessible databases where the general public can cross-check the information. Although blockchain technology is gaining popularity, tools like OpenTimestamps, which uses the Bitcoin blockchain for timestamp validation, and other Merkle Tree-based databases, like the Seal project that utilizes domain creation and Let’s Encrypt’s certified key, have been used in the past for similar purposes. These initiatives record certification requests in the Certification Transparency Log database.
TLDR: Reuters and Canon are working together through the Content Authenticity Initiative (CAI) to enhance the credibility of news images. Canon’s cameras embed digital signatures, GPS location, and metadata into the images, which are then sent to Reuters’ database and registered on blockchain platforms. When images are received, Reuters records the editing processes internally and confirmation hashes externally. Challenges include time-consuming digital signature computations and the detachment of edited images from their originals. The use of software from FotoWare can further ensure accurate image editing. Publicly accessible databases enable cross-verification, and previous initiatives have used blockchain technology and Merkle Tree-based databases.
Leave a Comment