Algorithms drive harmful eating broadcasts that lead to creator deaths in China

What happened

Eating broadcast creators in China have died or suffered severe harm after producing extreme eating content. A 24-year-old creator known as 干饭莹莹 (Chen Ziyi) with one million followers died; another known as 泡泡龙 weighed 320 pounds and died of heart failure; a third known as 三千哥 engaged in alcohol abuse challenges.

Social media platforms' algorithms promote extreme eating content that violates normal dietary practices and physical limits. Content that is shocking or extreme in nature—such as abnormally large food intake despite small body size, pushing physiological limits, or breaking dietary taboos—receives algorithmic amplification as a 'traffic code.'

Reported impact

The extreme eating broadcasts have resulted in deaths and serious health damage to content creators who engage in the practice. Creators transparently documented that their bodies were being treated as disposable commodities for generating engagement and revenue.

Index analysis

Platforms, MCN agencies, and creators form a chain of profit-seeking that packages life-threatening behavior as viral content. Regulatory enforcement has repeatedly failed, with the industry falling into a cycle of enforcement followed by brief compliance and then renewed violations.

Governance controls

China's anti-food waste law explicitly bans the creation, publication, and distribution of videos showing binge eating and food waste. Platforms have conducted multiple enforcement campaigns, yet creators evade rules through late-night hidden broadcasting, edited fake compilations, and disguised exaggerated eating.

Classification

Organizations

social media platforms (deployer), 干饭莹莹 (Chen Ziyi) (victim), Chinese monitoring bodies (regulator)

Category

Safety and wellbeing

Industry

Media, social platforms and marketing

Country

China

Harm status

Actual harm

Intent

Unintentional

Timing

Post-deployment

Harm types

Physical death, Physical injury, Human or fundamental rights

Harmed

Workers

Failure modes

Unsafe recommendation, Content moderation failure

AI systems

social media recommendation algorithms

Sources and evidence