Study: AI Personalized Recommendations Add 47 Minutes to Daily Screen Time

Tsinghua University's School of Journalism and Communication released the "2027 China Attention Economy Research Report" on January 5. Based on behavioral tracking data from 24,000 internet users, the report reveals AI personalized recommendations' profound impact on user attention.

The study found that Chinese internet users' average daily screen time reached 7 hours 23 minutes in 2027, a 47-minute increase from 2026. The largest increase was among adolescents aged 12-18, with daily screen time rising from 6 hours 12 minutes to 7 hours 20 minutes—a 68-minute increase.

The report identifies AI recommendation algorithm upgrades as the primary driver of increased screen time. New-generation recommendation algorithms can not only precisely predict user content preferences but also forecast users' emotional states and attention fatigue curves, delivering the most engaging content at optimal moments.

"This is essentially attention harvesting," said lead author Professor Peng Lan of Tsinghua University. "AI systems' objective function is maximizing user dwell time, not maximizing user wellbeing. These two objectives are fundamentally in conflict."

The report recommends that platforms incorporate "attention health" metrics into algorithm optimization objectives and provide users with proactive daily usage time management tools.