We need a credibility indicator marketplace to fight disinformation
To counter disinformation and revalue quality journalism we need a competitive ecosystem of credibility indexes to stimulate innovation and avoid a Ministry of Truth, de facto or otherwise.
The first thing I liked about the EU Commission’s High Level Expert Group (HLG) on Fake News was that — from Page One — they dismissed the term ‘Fake News’ as unhelpful and redefined their mission to focus on disinformation. The EC (a year after everyone else) got the memo, publishing its Communication (“Tackling online disinformation: a European Approach”) a few weeks later.
Brussels Bubble reports are often as verbose as they are functionally useless. Thankfully, the HLG report wasn’t. Reflecting perspectives from many stakeholders, it inevitably doesn’t go as far in many directions as some would like, but covers a lot of ground.
But while the HLG report touches upon the vital role which multiple Credibility Indicators must play in detoxifying the information sphere, the approach set out in the Communication risks creating a monopoly. And while it probably won’t, the perception will do just as much damage to fact-checking.
Source transparency: process over quality
The approach to Indicators is a good example of the HLG report’s balancing act. While researchers build machine learning systems to spot disinformation, others see using algorithms to judge content quality like putting robots in charge of the Ministry of Truth.
using algorithms to judge content quality like putting robots in charge of the Ministry of Truth
The HLG report therefore focuses on the use by social platforms of source transparency indicators which (my emphasis):
“allow users to assess where content … comes from and … verify the source, its ownership, and its adherence to ethical and journalistic codes. … Such indicators should have nothing to do with the quality of the content but rather with processes… Platforms should integrate these source transparency indicators … to better signal trustworthy and identifiable sources”
Gaming the Ministry of Trust
TL;DR: Victor Orban influencing what is considered ‘true’ online, and/or destroying the credibility of Europe’s best fact-checkers by tarring them with the EU flag.
The Communication, on the other hand, sets as a key principle the development of a much wider class of indicators - of content trustworthiness, “based on objective criteria and endorsed by news media associations… notably with the help of trusted flaggers”.
The emphasis on fact-checking continues by pledging support for an independent network of fact-checkers using an EC-built online platform.
I’m sure this is not the intention, but a Commission-supported network of fact-checkers, using Commission-endorsed methods and a Commission-supplied, EU-wide information system risks creating a single, EU-wide network of ‘approved’ fact-checkers.
Moreover, the temptation to aggregate their work into a single ‘approved’ set of Content Trustworthiness Indicators, and eventually use it to train an AI algorithm for assessing trustworthiness automatically, may be irresistible.
Even if it doesn’t, the risk is that it will be perceived as such. This could diminish the involved fact-checkers’ impact by opening their flank to accusations of being part of the Establishment.
a single Indicator … another algorithm to be gamed, or a de facto Ministry of Truth?
Because those being fact-checked will not take it lying down, particularly if Indicators are imposed via Regulation. They won’t be alone in asking: what happens when a single Indicator-set establishes credibility for the search and social platforms which determine the content billions see every day? Another algorithm to be gamed? A de facto Ministry of Truth? Both?
The Communication (rightly) sets out that members of this exclusive fact-checking club would have to meet high standards. But if this network defines truth online, its membership and methods will become political fast. Politicians will find a way to have their say. And as the HLP report says:
“not all European politicians and public authorities share the same level of respect for media freedom”
A market of competing indicators could solve this
The HLG report touches on this, albeit only in the context of:
- ‘user empowerment systems’, which “should give the user the opportunity to have content displayed according to quality signals… Actions should be taken to foster the creation of a competitive market of such applications”
- ‘increase transparency and efficiency of fact-checking’, with a “final goal … an open market for fact-checking that avoids a ‘monopoly of truth’”
I explored where this could lead earlier (left). While a War of the Credibility Indexes could poison the market for user-facing technologies (browser plug-ins, etc.), the alternatives are probably worse: all major platforms harvesting and onselling our attention using the same, centrally-approved indicators, or — fearing a user backlash — using none at all.
Either way, when people realise that platforms are influenced by a centrally-supported list of fact-checkers, expect a new SEO-like acronym: FCO (Fact Checking Optimisation).
expect a new acronym: FCO (Fact Checking Optimisation)
Competitive Innovation
Once platforms need — for whatever reason — to use external indicators to improve the quality of the content they show users, a competitive market is the best way to create them.
Source transparency indices like the Trust Project’s are the obvious first products reaching this market. They may be enough, but all systems get gamed eventually, so constant innovation will probably be necessary.
An “indicator marketplace” would support that innovation, allowing organisations to develop and deploy a variety of indicators for platforms to adopt, using techniques as diverse as AI, fact-checking and crowdsourcing.
another level of innovation as platforms combine different indicators in new ways
Crucially, platforms could then use some or all of them, adding another level of innovation as they experiment with combining different indicators and tweaking user interfaces to produce the best possible user experience, and constantly changing the mix to stay ahead of those trying to game them.
Such a marketplace must embody transparency
All search and social platforms, for a start, would have to be transparent about the indicators they’re using. Both HLG and Communication seem to support this, while respecting the platforms’ right to keep their algorithms secret.
Each product, in turn, would have to be as transparent as possible about its sources and methods, and should be open to critique and user engagement.
transfer arguments off the platforms, where they amplify the reach of contentious content
The Credibility Index Providers explored previously, for example, explicitly cite the sources used by each fact-checker they aggregate, and allow users to discuss and contest the aggregated score. This would transfer arguments over content credibility away from the platforms, where arguments amplify contentious content, analogous to the way Wikipedians fiercely argue on discussion pages most people never visit.
But there are probably dozens of other ways to create good Indicators, and as long as they are transparent they should be in the market, fighting for the platforms’ business.
Multilingual and multicultural
Such a marketplace must also be multilingual and multicultural.
The English-speaking world has been the first to suffer from disinformation, and is home to the world’s dominant online technology companies. Many of the early solutions are therefore coming from Americans and Brits. As a barely bilingual Australian in Brussels, I fear the size of the linguistic blind spots I cannot see.
One size, in other words, will not fit all. A variety of innovations must be tested across Europe’s diverse media landscape to establish a competitive, multicultural marketplace for credibility indicators, and avoiding the creation — or even the perception — of EU-approved Indicators, forced on the platforms by EU Regulation.
If you found this interesting, your Applause will help others find it.
My Hub curates 100s of annotated resources about fake news, facebook, social media, filter bubbles and more. Subscribe to get my next posts, plus the best of all the Stuff I Curate, in your Inbox, or get just the High3lights via my CuratorBot. It can also put us in touch, or we can connect via my awful brochureware site.
