Platform‑level content exposure datasets

Researchers and platforms should create shared, privacy‑preserving datasets that record adolescents' exposure to specific content types (e.g., self‑harm posts, harassment, prosocial resources) with standardized timestamps and demographic covariates so causal links to outcomes can be tested. These datasets would allow reproducible analysis of heterogeneity (who is harmed, helped, or unaffected) and of dose–response relationships across real use patterns. — If platforms and scientists adopt shared, standardized exposure datasets, policy can move from arguing about 'time on screens' to regulating concrete, verifiable content harms and protections.