Join The Community | A Private Members Club for the Tech Economy
Learn more
Apr 6, 2022 7 min read

Facebook Knew It Was Promoting Fake News, But Couldn't Stop It

Wiser! Essay: Facebook's algorithms were meant to "down rank" misinformation to stop it from spreading. Instead, the AI did the opposite. And promoted harmful content by as much as 30%. It took Facebook engineers 6 months to fix it!


w/Promotion

The Easiest Tool for Writing Twitter Threads in the World (IMHO)


When Facebook's "DownRanking" Algorithm Didn't Work

BackStory:  Repeat misinformation offenders saw their views spike by as much as 30% due to a flaw in the Facebook algorithms

The Verge has reported that Facebook engineers identified a fault in an AI system called "downranking".

Downranking does exactly what it says on the tin! Downranking is when repeat fake news offenders, as flagged by Facebook's army of content moderators when it comes to posting harmful content, are slowed down and suppressed.

But instead, the Facebook algorithms accelerated their posts in the News Feed. By as much as 30% according to a statement to The Verge from Facebook.

One In 3 Internet Users Fail to Question Misinformation

According to the latest data from the UK regulator, Ofcom, 1 in 3 Internet users fail to question online content.

Read the full story.

This isn't a paywall. Join for free and get access to all member content.

Subscribe
Already have an account? Sign in Sign in
Great! You’ve successfully signed up.
Welcome back! You've successfully signed in.
You've successfully subscribed to Wiser! - Business & Technology Explained.
Your link has expired.
Success! Check your email for magic link to sign-in.
Success! Your billing info has been updated.
Your billing was not updated.