Fujitsu and Japanese Uni propose 'endorsement layer' to make the internet trustable
Hope to combat misinformation with graph of observations that prove what you’re reading is reliable
Fujitsu and Japan's Keio University have proposed adding an "endorsement layer" to the internet, to verify information and prevent the flow of fake news and disinformation.
The video details a scenario in which a social media user posts an image of what appears to be unusual quantities of dirty brown water coursing over a dam's spillway and into a river with an opinion that a flood is imminent. The scenario assumes the post is taken to be true – sparking anxiety among residents, possible panic buying, or even an inappropriate emergency services response.
The endorsement layer the authors envisage would scour the internet for authoritative data sources, such as water level sensors. Users who view a post about a possible flood could request to access such data, which would be published by the endorsement layer. A better-informed reader would therefore know not to raid supermarkets for a year's worth of toilet paper, and emergency services would save their resources for more pressing tasks.
The proposal imagines the endorsement layer would not rely on a single sensor or information source, but instead offer users the chance to add data to the endorsement layer. The result would theoretically be "an endorsement graph with a data structure expressing the connection between additional information linked to the data."
- W3C's planned transition to HTTPS stymied by legacy laggards
- W3C overrules objections by Google, Mozilla to decentralized identifier spec
- FIDO Alliance says it has finally killed the password
- Japanese space agency to put massive HPC cloud to the test
The content and recommendations drawn from of the endorsement graph would be overlaid on the web, or apps, so that users could understand the context of whatever they're looking at. Browser extensions that allow users to filter info from the endorsement graph to meet their needs are suggested as a possible refinement.
"This is expected to make it possible to prevent the use of unreliable data and the re-spread of disinformation and fake news," the white paper states, adding "Furthermore, by expanding the scope of judgment, the scope of economic and social activities will be expanded, and the optimal method will be realized in society."
Which is perhaps a little too optimistic, but the prospect of the "trustable internet" is not. The authors note that the World Wide Web Consortium has already proposed a verifiable credentials data model, the EU is working on a Digital Identity Architecture and Reference Framework, and the Linux Foundation's Hyperledger Indy could help to establish and verify the identity of contributors to the endorsement graph.
The idea of a trustable internet is of course, years away from becoming reality. And the paper doesn't address the issue of how to get netizens to trust the trustable internet – surely a significant concern when conspiratorial thinking is prominent and trust in institutions low. ®