← the record
AIAAIC-0741

Facebook data leak exposes Balkan troll farm political disinformation

Troll farms reached 140 million US Facebook users per month and nearly half of all Americans in the run-up to the 2020 presidential election , according to a leaked document. Authored by former Facebook data scientist Jeff Allen and leaked to the Technology Review , the document revealed that 75 percent of US users consumed troll content not because they followed a page but because Facebook’s recommendation engine proactively served it to them. Largely based in the Balkans, the troll farms were seen to have targeted four primary groups - American Indians, Black Americans, Christian Americans, and American women - with a mixture of propaganda, misinformation and disinformation. According to the Technology Review , Allen left Facebook shortly after writing the document, partly because the company 'effectively ignored' his research. System 🤖 Recommendation system Developer: Meta/Faceboo k Country: US Sector: Politics Purpose: Scare/confuse/destabilise Technology: Content recommendation system Issue: Accuracy/reliability; Mis/disinformation Resource s 📃 Facebook (2019). How Communities are Exploited on our Platform: A Final Look at the 'Troll Farm' Pages

Date it happened
2021-09-01
Organisation involved
Facebook
Product, system or model
Facebook recommendation system
Where this came from
Share this incident
XLinkedInFacebookWhatsAppEmail
Attribution

This incident was imported from AIAAIC and is used under CC BY-SA 4.0. Our additions to it — the structured fields, the translation, the checks against other reports — are published under the same licence.

This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.