Fixing the science of digital technology harms.

When a London coroner's inquest found that 14-year-old Molly Russell died by suicide in 2017 due to "depression and the negative effects of online content" (1), the decision posed a challenge for the science of digital technologies. The investigating psychiatrist reported losing sleep for weeks just from reviewing disturbing material to which digital platforms had exposed Molly. Yet scientists are debating whether and how such technologies pose harm to individuals and society. From social media to artificial intelligence (AI), the accelerating development and deployment of new digital technologies have outpaced the capacity of science—constrained by limited resourcing and restrictions on proprietary data—to effectively assess their impacts. We must urgently fix this science-policy ecosystem to support timely, measured assessments of sociotechnical risks that can lead to real change. We must transform how science and policy is organized, better manage the risk of being wrong, and develop policies that synchronize science and innovation. [ABSTRACT FROM AUTHOR]