TikTok FYP Algorithm Ignores Negative User Feedback

Northwestern University computer scientists have found that TikTok users have less agency over their For You Page (FYP) content than they might believe. The study, published in a recent paper, investigated user concerns that the platform's algorithm does not effectively incorporate negative feedback, leading to the continued appearance of unwanted videos on their FYP. The research team, specializing in "algorithm audits," aimed to understand how online platforms function and fail.

According to co-author Piotr Sapiezynski, the study revealed that while engagement signals, including explicit feedback like "not interested" or "see less," do influence the FYP, their impact is temporary. The algorithm tends to revert to its previous patterns unless users consistently provide the same negative feedback over an extended period. This suggests that the "not interested" feature, while available to users, may not provide the lasting control over content selection that users expect.

Sapiezynski noted the apparent contradiction of platforms offering such feedback mechanisms if they are not consistently effective. The research group's work focuses on auditing algorithms to identify how they operate, where they falter, and the potential harm they can cause to individuals and society. The findings specifically address anecdotal reports from TikTok users who felt their attempts to curate their feed through negative signals were largely ignored by the platform's powerful recommendation engine.