Skip to content
Submit incident
AII-1002-54703
AllegedThe record is a lawsuit, complaint or accusation that has not been decided.

YouTube's algorithm allegedly amplifies eating disorder content to adolescent girls

YouTube's recommendation algorithm reportedly escalated eating disorder content to teenage girls who engaged with weight loss videos, according to research by the Center for Countering Digital Hate. One 14-year-old girl whose feed was filled with such content was subsequently diagnosed with anorexia.

March 2020 Β· YouTube Β· Safety and wellbeing
March 2020
AII-1002-54703View source β†—
Quick facts
Category
Safety and wellbeing
Country
United States
Source
cbsnews.com
Date
March 2020
Drafted by
Raidu Incident Detection

What happened

In spring 2020, 14-year-old Anna Mockel began watching weight loss and diet videos on YouTube while housebound during COVID lockdown. As she watched, the platform's 'Up Next' recommendation panel shifted from videos featuring thin girls to how-to weight loss content, then to videos glorifying skeletal bodies and extreme dieting hacks, with recommendations for 500-calorie daily diets.

Anna copied the content she saw, restricted her diet severely, and was subsequently diagnosed with anorexia. She underwent two hospitalizations and spent three months at a residential treatment center, beginning recovery at age 16.

A report released by the Center for Countering Digital Hate found that when YouTube users show interest in diet and weight loss, nearly 70% of videos the algorithm pushes recommend content likely to worsen or create body image anxieties. These videos average 344,000 views each and carry ads from major brands including Nike, T-Mobile and Grammarly.

Reported impact

Anna states that YouTube perpetuated her eating disorder by creating a competitive online community that reinforced her condition and prevented her from recognizing it as a problem.

Classification

Category
Safety and wellbeing
Industry
Media, social platforms and marketing
Country
United States
Harm status
Actual harm
Intent
Unknown
Timing
Post-deployment
Harm types
Psychological
Harmed
Children
Failure modes
Biased model, Unsafe recommendation
AI systems
YouTube recommendation algorithm

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

cbsnews.com
YouTube's algorithm allegedly amplifies eating disorder content to adolescent girls