A TrustLab Report

Study on the Role and Effects of the Use of Algorithmic Amplification to Spread Terrorist, Violent Extremist and Borderline Content

Between July 2022 and May 2023, and now released in November of 2023, TrustLab conducted a study to examine how algorithms across major social media platforms proactively present online content to users and in particular how their algorithms amplify Terrorist, Violent Extremist and Borderline Content to users. The study analyses Facebook, Instagram, TikTok, Twitter, and YouTube covering 8 markets/languages across the EU.

The measurement study was commissioned by the European Commission, Directorate-General for Migration and Home Affairs for the EU Internet Forum.

The primary questions that the study addresses are:

Findability: How easy is it to find harmful content, measured in the count of harmful posts during a one-hour interval?
Removal Rate: What is the amount of harmful content removed by the platform?
Removal Time: How much time did it take for the harmful content to be removed?
User Sentiment: What is the public (crowdsourced) perception of the appropriateness of the harmful content, measured on a 5-point Likert scale?
Amplification: What is the amount of harmful content in a user's feed, measured as a percentage of the first 30 posts in the feed?

TrustLab was selected by the European Commission to develop measurement methodology and conduct post-measurement analysis in collaboration with Professor Theodoros Evgeniou, Professor of Decision Sciences and Technology at INSEAD who researched the risk of radicalisastion linked to the dissemination of TVE content.

Specific findings include:

Evidence of amplification:

The study validates the hypothesis that increased interaction with TVE content and Borderline content results in higher amplification of such content to users.
All platforms showed amplification of TVE-promoting content in their feeds.
Twitter showed the highest level of amplification, YouTube ranked second highest, while TikTok showed the least.

Findability:

Twitter has the highest findability scores of TVE content across the five platforms, while YouTube had the lowest findability scores.
Italian language content had the highest findability scores.
Violent Left-Wing content had a significantly higher Findability score than other TVE types.

Removal Rates:

More than 90% of TVE content found by TrustLab remained on the platform by the end of the evaluation period 8 weeks later.
TikTok removed more content than other platforms and Twitter removed the least.