Do short-form video algorithms actually increase how much we use our phones?
Yes, and the effect is large and causal. A 2025 study of 1,764 US mobile users compared people who adopted TikTok (which was then uniquely short-form video) to those who adopted Instagram or Facebook during the same period. TikTok adopters increased their total mobile usage by 17% — about 28 extra minutes per day — and reduced their time away from the phone by 20% (29 fewer minutes away per day) [1][2]. The effect was strongest in people who had previously used their phones less, meaning the algorithms hook lighter users the most. Surprisingly, older adults showed larger increases in the number of daily phone sessions than younger users, suggesting the design features are broadly addictive across age groups [1][2].
A separate study of 2,646 Chinese college students found that the personalized recommendation algorithm directly predicted addiction tendency, and this effect was partly explained by the psychological state of 'flow' — the immersive, time-lost feeling users get when swiping through tailored content [4]. The same study showed that people with lower self-control were especially vulnerable to this algorithmic pull [4]. Together, these findings show that the recommendation engine is not a neutral tool; it is a key driver of compulsive use.
What are the real-world harms of this addiction?
The most well-documented harm is to sleep. In a study of 1,629 Chinese adolescents, 31% had poor sleep quality, and short-form video addiction was directly linked to worse sleep scores. The effect was partly mediated by social anxiety — meaning addiction made teens more socially anxious, which in turn disrupted their sleep [3]. The direct effect of addiction on sleep quality accounted for 62% of the total harm, while the indirect path through social anxiety accounted for 38% [3].
Learning and motivation also suffer. A survey of 517 Chinese vocational college students found that addiction to short videos reduced both intrinsic and extrinsic learning motivation, and indirectly lowered students' sense of learning well-being [5]. The immersive 'flow' experience that makes these videos so engaging was actually the gateway to addiction and its negative effects on motivation [5].
Mental health risks are serious. A 2025 review paper notes that algorithmic curation, endless scrolling, and feedback loops of social validation can trigger social comparison, anxiety, and depressive symptoms in teenagers [7]. Another study developed an AI model to predict which short-form videos might trigger suicidal thoughts in viewers, finding that certain content themes are reliably linked to suicide risk [8]. A 2026 commentary in the Journal of the Pakistan Medical Association warns that the constant dopamine-driven reward cycle leads to attention deficits, emotional numbness, and even 'passive nihilism' — a loss of meaning and engagement with life [10].
Can the platforms be redesigned to reduce these risks?
The evidence suggests that specific design features are the root cause, and changing them could help. A 2024 interview study found that users are frustrated by the lack of transparency in content reporting, the difficulty of training algorithms to avoid distressing content, and the limitations of post-exposure feedback (like 'not interested') [9]. The researchers proposed three design fixes: making the reporting process more transparent, giving users granular control over what the algorithm recommends, and allowing preemptive adjustments to content feeds [9].
However, a 2026 experiment on 180 children found that while swipe interfaces and recommendation algorithms increased the rate of video and category switching, these features did not directly impair short-term or working memory in a single session [6]. This suggests that the cognitive harms may accumulate over time with repeated use rather than appearing after one exposure. The same study noted that higher combined levels of video and category switching were marginally associated with worse working memory performance, hinting that the rapid context-switching design may still be harmful in the long run [6].
About These Sources
This answer is built on 10 peer-reviewed studies — published from 2022 to 2026, 9 from 2024 or later, 4 in Q1 journals, collectively cited 203 times — selected as the most relevant from 11 studies that passed quality screening, drawn from 48 papers retrieved from a database of over 500 million.
Sources used in this answer
Short-form video platforms drive mobile usage
In a study of 1,764 US mobile users, adopting TikTok increased total daily phone use by 17% (about 28 minutes) and reduced time away from the phone by 20% (29 fewer minutes away per day), with larger effects on lighter users and older adults [1].
Short-form video platforms drive mobile overuse
Same study as [1] — confirms that short-form video platforms causally drive mobile overuse, increasing usage duration and reducing time away from the phone, especially during daytime hours [2].
Adolescents’ short-form video addiction and sleep quality: the mediating role of social anxiety
Among 1,629 Chinese adolescents, 31% had poor sleep quality; short-form video addiction directly predicted worse sleep (62% of the effect) and indirectly through increased social anxiety (38% of the effect) [3].
Personalized recommendation and short-form video addiction tendency: the mediating effect of flow and the moderating effect of self-control
In 2,646 Chinese college students, personalized recommendation algorithms directly predicted addiction tendency, mediated by the immersive 'flow' state, and moderated by self-control (lower self-control increased vulnerability) [4].
Effects of Short Video Addiction on the Motivation and Well-Being of Chinese Vocational College Students
Among 517 Chinese vocational college students, short-form video addiction reduced both intrinsic and extrinsic learning motivation and indirectly lowered learning well-being, with flow experience acting as a gateway to addiction [5].
Endless Swipes and Recommendations: The Impact of Short-Form Video Platforms on Context-Switching and Children's Working Memory.
In a between-subjects experiment with 180 children, swipe interfaces and recommendation algorithms increased video and category switching but did not directly impair short-term or working memory in a single session; higher combined switching was marginally associated with worse working memory [6].
Exploring the Correlation Between Teenagers Short-Form Video Consumption and Symptoms of Anxiety and Depression
A review paper finds that algorithmic curation, endless scrolling, and feedback validation loops in short-form video platforms can trigger social comparison, anxiety, and depressive symptoms in teenagers, especially those with pre-existing vulnerabilities [7].
Short-Form Videos and Mental Health: A Knowledge-Guided Neural Topic Model
A novel AI model that integrates medical knowledge on suicide risk factors with user-generated content can predict which short-form videos may induce suicidal thoughts in viewers at the time of upload, outperforming standard machine learning benchmarks [8].
Exploring Design Approaches for Reducing Viewers' Discomfort with Distressing Short-form Videos
Through in-depth interviews, users reported frustration with lack of transparency in content reporting, difficulty tailoring algorithms to avoid distressing content, and limitations of post-exposure feedback; three design approaches were proposed to reduce discomfort [10].
The dopamine trap: Effects of short-form video (SFV) addiction on mental health
A commentary argues that excessive short-form video consumption creates a dopamine-driven addiction cycle leading to attention deficits, emotional numbness, and passive nihilism, and urges healthcare providers to screen for SFV consumption during adolescent mental health assessments [11].