Study finds TikTok's algorithm requires repeated negative feedback to permanently remove unwanted content
Northwestern University researchers tested TikTok's FYP algorithm and found that negative feedback (like swiping away or marking "not interested") only temporarily affects recommendations. The algorithm gradually reverts to earlier suggestions unless users repeatedly provide consistent negative signals. This suggests users have less control over their feed than they might believe, as the system requires sustained, repeated feedback to permanently remove unwanted content.
Key quotes
TikTok's FYP responds to negative feedback only temporarily
repeated consistent signals are needed to keep unwanted videos from reappearing
Engagement signals affected recommendations briefly, but the system gradually returned to earlier suggestions unless users repeatedly provided the same negative feedback
From the article
You might also wanna read
TikTok users don't have as much agency over their FYPs as they think
TikTok Tests User Control Feature for AI-Generated Content Visibility
TikTok is testing a new AI content control feature that allows users to manage how much AI-generated content appears in their video feed. Th
The Verge·7mo ago
How TikTok Industrialized Human Attention and Transformed Media Consumption
The article examines how TikTok has fundamentally transformed media consumption and culture by industrializing human attention through short
How TikTok’s algorithm is affecting you, from the content you see to what you can do to regain control
Why platforms should let users filter out AI-generated content
The article argues that while major platforms like YouTube, Instagram, and TikTok have implemented AI content labeling systems, they have fa
The Verge·1mo agoWhy platforms should let users filter out AI-generated content
The article argues that while major platforms like YouTube, Instagram, and TikTok have implemented AI content labeling systems, they have fa
theverge.com·1mo ago
Whistleblowers: Meta and TikTok Allowed Harmful Content After Research Showed Outrage Drives Engagement
Whistleblowers reveal that social media companies Meta and TikTok knowingly allowed more harmful content on their platforms after internal r
bbc.com·3mo ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.