Digital technologies rewire childhood learning and social life: well-designed AI tools can boost skills and access, but uneven access, attention harms and mental-health risks mean benefits are mixed and policy must balance opportunities with protections.

Main Finding

Digital technologies are increasingly embedded in children’s lives and are reshaping learning, social relationships, identity formation, cognition, and health. These effects are complex and mixed: technologies can enhance access to learning and new skills while also disrupting traditional educational processes and producing unintended social, mental-health, and equity consequences. Effective responses require balanced engagement, stronger digital literacy, and evidence-informed policy.

Key Points

• Rising exposure: Children and adolescents worldwide are spending more time on digital devices and platforms, both for education and social interaction.
• Dual effects on learning: Digital tools can improve personalized learning, access to information, and skill acquisition, but may also reduce attention, deepen superficial learning, and weaken some traditional pedagogical benefits if poorly integrated.
• Social and identity impacts: Online interactions reconfigure peer relationships, social norms, and identity construction (e.g., through curated self-presentation and constant feedback), which can have developmental consequences.
• Mental and physical health risks: Increased screen time and certain platform designs are associated with sleep disruption, attention issues, anxiety, and depressive symptoms for some children, with heterogeneous effects across individuals.
• Inequality and access: Benefits are uneven—socioeconomic status, digital infrastructure, parental support, and school capacity shape whether children gain from digital tools or experience harms.
• Rapid change and uncertainty: Technology evolves faster than research and policy, creating uncertainty about long-run outcomes and causal pathways.
• Policy levers: Promote balanced use, strengthen digital literacy curricula, regulate design and data practices that affect minors, and support evidence generation (evaluations and longitudinal monitoring).

Data & Methods

• Nature of evidence: The article synthesizes findings from a mix of cross-sectional and longitudinal observational studies, randomized controlled trials (RCTs) in educational settings, qualitative research, and natural experiments. It also draws on theoretical and conceptual work about development and technology.
• Common data sources cited:
  • Time-use surveys and parental reports of screen time
  • Educational outcome datasets (test scores, attainment)
  • Cohort studies linking media exposure to health/behavioral outcomes
  • Platform or app usage logs (digital trace data) where available
• Methodological strengths:
  • Longitudinal cohorts help track developmental trajectories
  • RCTs provide causal evidence for specific digital interventions in education
• Methodological limitations and gaps:
  • Many studies are correlational, making causality difficult to establish
  • Rapid technological change makes older studies less applicable
  • Heterogeneous measures of “screen time” and outcomes hinder synthesis
  • Underrepresentation of low- and middle-income contexts and of marginalized groups
  • Limited access to proprietary platform data constrains understanding of algorithmic effects
• Recommended research improvements:
  • More preregistered RCTs of educational technologies
  • Long-term cohort studies combining surveys and digital trace data
  • Improved, standardized outcome measures (cognition, socioemotional outcomes)
  • Policy experiments and evaluations (e.g., school-level rollouts, platform regulation)

Implications for AI Economics

• Human capital formation and productivity

  • Early and formative exposure to digital technologies shapes skill acquisition (technical, cognitive, socioemotional), altering the future workforce’s composition and productivity. AI-driven educational tools can accelerate skill formation but may also produce uneven returns if access is unequal.
  • Economists should model heterogenous treatment effects of digital/AI education tools across socioeconomic groups and over lifecycle horizons.

• Labor market and inequality dynamics

  • Differential access to high-quality digital learning resources may widen skill and income gaps. Policymakers and firms investing in AI-based educational products risk reinforcing inequality unless targeted distribution and subsidies are used.
  • Consider the potential for AI-enabled tutoring and credentialing to revalue certain skills, affecting demand for labor and wage structures.

• Market structure and competition

  • Platform firms that control large amounts of child- and youth-oriented data may gain market power in educational and entertainment markets. This raises competition-policy issues (entry barriers, bundling of services, data advantages).
  • Regulation of data collection and minors’ privacy will shape market incentives for AI product design and monetization strategies.

• Externalities, public goods, and policy design

  • Negative externalities (e.g., mental-health costs, attention harms) create roles for public intervention—school curricula, regulation of algorithmic content, or limits on certain attention-capturing features.
  • Positive externalities (broader access to learning) justify public investment in digital infrastructure, teacher training, and evidence-based AI tools for schools.

• Data, measurement, and evaluation for policy

  • Economists should leverage administrative educational data, randomized rollouts, and digital trace data (with strong privacy protections) to estimate causal effects and inform scalable policy.
  • Cost-effectiveness analyses of AI-enabled interventions versus traditional approaches will be crucial for budget-constrained education systems.

• Regulatory and ethical considerations affecting economic incentives

  • Privacy rules (e.g., limits on collecting minor data) and design standards (for engagement mechanics) will affect firms’ revenue models and incentives to develop child-focused AI products.
  • Policies that require transparency or limit targeted advertising to children will change market strategies and could reduce funding for some free services, with distributional consequences.

• Research agenda for AI economists

  • Estimate long-run returns to early digital/AI-based interventions on earnings and welfare.
  • Quantify distributional impacts and identify policies that maximize inclusive human capital development.
  • Study how platform design and algorithmic personalization influence behavior, learning, and market outcomes among youth.
  • Evaluate regulatory interventions (privacy, competition, product standards) using experimental or quasi-experimental designs.

Overall, the integration of digital technologies into childhood development profoundly interacts with economic processes—human capital formation, market structure, and policy design. AI economists should prioritize causal, longitudinal, and policy-relevant research to guide interventions that harness benefits while mitigating harms and inequality.