Archive — US — techcrunch
Study of TikTok, X ‘For You’ feeds in Germany finds far-right political biasahead of federal elections
A study by Global Witness reveals significant far-right political bias in algorithmic feeds on TikTok and X ahead of Germany's federal elections, with 78% of recommended political content on TikTok favoring the far-right AfD party. The research highlights how social media algorithms amplify right-leaning content, raising concerns over transparency and democratic implications.
Key points
- Global Witness analyzed algorithmic recommendations for political content on TikTok and X, finding a bias towards far-right AfD.
- On TikTok, 78% of the recommended content for non-partisan users supported the AfD, compared to 64% on X.
- The research demonstrated that users were exposed to right-leaning content over twice as much as left-leaning content in the lead-up to elections.
- TikTok and X's algorithms likely prioritize engagement, which may inadvertently amplify political bias.
- The findings are consistent with previous studies indicating right-wing bias in social media algorithms.
Why it matters
- The study's findings echo concerns raised in similar research on social media's role in political elections in the U.S., Ireland, and Romania.
- Historically, algorithmic biases have led to significant political ramifications, influencing electoral outcomes through distorted information visibility.
- The EU's Digital Services Act aims to improve transparency around algorithmic systems and their potential biases, but full implementation is still pending.
The documented bias in social media algorithms presents serious implications for democracy and electoral integrity, highlighting an urgent need for transparent algorithm practices and stricter regulatory measures.