X and TikTok algorithms push pro-AfD content to non-partisan German users
On TikTok, 78% of algorithmically recommended political content (from unfollowed accounts) was pro-AfD; on X, 64%, far exceeding the party's ~20% polling.
Executive summary
AI-generatedThis summary was generated by AI from the original report to make it easier to scan and cite. It is not a substitute for the source — read the original above.
Ahead of Germany's federal elections, Global Witness set up nine new social media accounts — three each on TikTok, X, and Instagram — designed to signal interest in German politics without favoring any particular party, in order to test what political content each platform's recommendation algorithm would surface to a neutral new user.
On TikTok, 78% of algorithmically recommended political content from accounts the test users did not follow supported the far-right Alternative für Deutschland (AfD); on X, the figure was 64%. Both figures far outstripped the roughly 20% support the AfD was drawing in contemporaneous polling. More broadly, non-partisan users were shown right-leaning content more than twice as often as left-leaning content in the run-up to the vote.
Global Witness concluded that TikTok's and X's recommendation systems were systematically over-amplifying AfD-aligned content relative to the party's actual public support. TikTok disputed that the small-scale test reflected typical user experience, while X did not respond to the findings.
Think this summary is wrong? Contact us.