Twid: fighting desinformation on Twitter with authenticity

Experiment to see if adding authenticity to Twitter messages (using IRMA signatures) can counter the spread of disinformation.

With the rapid growth of online information came the exponential spread of disinformation. Recent examples include fake news about Covid-19 and false political advertisements. Disinformation is often counteracted by fact-checking. However, it is difficult and labor-intensive to rectify and remove false information that has already been published. Furthermore, discussions about what is true or not, certainly in a political or morally charged setting with many perspectives, quickly derail. An alternative approach, as proposed by one of the applicants, is to concentrate on authenticity of information instead of on truth. This Twid project applies this to Twitter. It provides Twitter users – with Twid’s browser plugin – with certainty about who is the source of a Twitter message known as tweet (“source authenticity” of tweets) as well as certainty that a message has not been modified or changed since its publication as tweet (“message authenticity”). The proposed approach is close to traditional approaches, where, for instance, knowing that a message comes from a certain newspaper helps people in their credibility assessment. Unfortunately, on social media platforms, such valuable background information about the source of messages is often lacking. The Twid project aims to fill this gap in the context of Twitter, by providing users with verified information about the origin of tweets. Concrete, the project proposes a browser-plugin that allows Twitter-users to link personal verified information to their accounts and sign their tweets. Other users who have the plugin enabled will then be able to see this information about who has posted/signed the tweets, helping them with judging the credibility of the messages. (For instance, knowing that someone is working at a renowned hospital might give credibility to their tweet about COVID-19). To link personal verified information with Twitter, the Twid project utilizes the existing identity management app IRMA, which already allows users to prove properties about themselves and sign digital content. With Twid, users can link the personal information that they have collected in IRMA (for instance, their real name from the Dutch Civil Registry, their city of residence or that they have an email-address ending with, e.g., @radboudumc.nl) to their Twitter account and use it to sign their tweets. This means that source and message authenticity is guaranteed via digital (cryptographic) signatures. The Twid project aims to demonstrate the feasibility of this approach with a Proof of Concept (PoC). This PoC is meant to demonstrate the feasibility of guaranteeing authenticity to various message services. The ultimate goal is that Twitter adopts the technology. The PoC concentrates on Twitter since Twitter is very vulnerable to fake news and is itself also exploring options to fight it. The PoC plugin can in principle be used by many people (such as politicians and opinion leaders in The Netherlands, and beyond) and may thus attract much visibility and put pressure on Twitter. At the same time, the applicants are seeking contacts with Twitter about Twid.