Twitter algorithms boost right-wing news, commentary

Occurred: October 2021

Twitter discovered that its algorithms disproportionately boosted right-wing news and commentary content. 

Covering politicians’ tweets and political content from Canada, France, Germany, Japan, Spain, the UK and US from April 1 to August 2020, Twitter found that right-wing content was more amplified algorithmically relative to a reverse-chronological timeline in every country, other than Germany.

Twitter director of software engineering Rumman Chowdury and machine learning researcher Luca Belli admitted it is unclear why its algorithms behaved this way. 

'Further root cause analysis', they say, 'is required in order to determine what, if any, changes are required to reduce adverse impacts by [Twitter's] Home timeline algorithm'. 

Operator: Twitter
Developer: Twitter
Country: Canada; France; Germany; Japan; Spain; UK; USA
Sector: Politics
Purpose: Recommend content
Technology: Recommendation algorithm
Issue: Bias/discrimination - politics
Transparency: Governance; Black box; Complaints/appeals

Page info
Type: Incident
Published: October 2021