For the first time, big tech platforms have reported on their efforts against misinformation. A lot of hot air, but TikTok made an effort.
The EU is pouring money into the fight against fake news. Big tech contributes too, but they decide how we proceed, and therefore they make money on them at the same time. The very first reports are now ready from the big platforms that signed up to EU’s voluntary ‘Code of Practice on Disinformation‘.
Metas’ report is interesting. Meta itself believes that its efforts to put warning labels on fake news are effective. Indeed, according to them, these labels stop 95% of users from clicking on them. Well well, you think, that sounds pretty good, but when you look at so-called click-through rates on Facebook, it’s 2.5% for publicist media content and a maximum of 1.6% for advertising (most are under 1%). So a 5% click-through rate for fake news is really high.
Further, Meta reports that among those who click on the warning-marked fake news, 25% on Facebook and 38% on Instagram do not share it. WHAT? That means 75% and 62% respectively actually share fake news.
Clicks and content sharing equals money. Like Meta, other platforms like Youtube and Tiktok also make money from fake news. Fake news gets six times more clicks than factual news, according to Washington Post.
Unfortunately, we let the platforms decide how to inform citizens and users about disinformation. Metas 26 fact-checking organisations worldwide accept that once they have fact-checked something, Meta decides how users should be made aware of the fake news. Meta won’t remove them because risk violating freedom of expression, it says, so instead, it’s turning down their viral potential. Highlighting the fake over the truth is hopeless – especially in a digital world where there is now more fakes than facts.
It’s a victory for the EU that the reports from the 30 signatories of the code are now available, but they also show how hopeless the fight against fake news is, how much hot air, and how opaque the major platforms are because it’s hard – if not impossible – to actually measure their efforts.
TikTok is the one that has tried to make the most of it. In relation to Commitment 14 on ‘malicious actors’, TikTok writes that in Q3 2022 it removed over 800,000 fake accounts with over 18 million followers in Europe (10,000 were in Denmark with just over 260,000 followers) and that the fake accounts represent 0.6% of the total number of active users
Reports from other major platforms are patchy. Microsoft reports that in one month they have removed 850,000 fake accounts on LinkedIn (14,500 in Denmark). Meta reports that in Q3 it removed 1.5 billion fake accounts worldwide, but does not provide figures for EU countries. And Google is even vaguer: Over 10 million domains related to Google Search and 10,000 YouTube channels were related to malicious actors, and Google refers to a website where they list their endless removals. Rather useless reports and they’re not verified by independent third parties either.
The EU Commission acknowledges that the reports are only a first step and tells Politico that when the Digital Service Act, DSA, comes into force later this year, the EU will be more forceful.
However, we should review how we spend taxpayers’ money to fight fake news. It’s unwise to let profit rule. Danish Meta partner Tjekdet.dk also accepts that posts flagged as fake remain on Meta’s platforms. On the other hand, tjekdet.dk contains many good tips on how to spot fake news yourself, and there is no doubt that media literacy is an important weapon. However, it is far from enough.
The EU should demand that anything labelled false by recognised fact-checking organisations be removed, and then start handing out fines as soon as the DSA is in force. Fines of up to 6% of turnover for failing to combat misinformation effectively should have an impact.
Disinformation/fake news is a big topic in our upcoming report “Big Tech’s Soft Power in Denmark” – a project, supported by the Drejers Fond, the Borgerretsfonden, the Jyllands-Posten Fonden, the Politikens Fond and the Culture and Information Association. It will be published by DataEthics.eu in 2023. The idea is to look into how big tech supports academia, publicist media and think tanks. This applies both financially and in terms of creating narratives, formulating policies, etc. The aim is to equip Danish decision-makers to better understand how big tech works.
Any tips on cases and examples, do send us an email at firstname.lastname@example.org
This column was first published in Danish in the national daily Politiken