AI Trials Observatory launches: 93 RCTs on how AI affects mental health and beliefs
DG_Rand · x · 2026-11-11
A new living systematic review, AI Trials Observatory, aggregates randomized trial evidence on how generative AI affects people, starting with 93 trials across two outcomes.
Mental health (52 trials, 211 effects):
- Therapeutic AI vs. no-activity controls improves anxiety (+0.33), depression (+0.25), wellbeing (+0.27), and psychological distress (+0.36)
- Versus alternatives like human chat, gains shrink to +0.05, suggesting much of the benefit is having something to talk to
- General/companion AI shows +0.64 on anxiety, though most other outcomes lack enough studies
Belief accuracy (42 trials, 124 effects):
- Corrective AI reduces misinformation belief (+0.61) and conspiracy belief (+0.45)
- General AI improves factual misperceptions (+0.65) but only +0.10 on misinformation belief, even -0.13 vs. a fact sheet
Effect sizes are Hedges' g; averages over 10+ studies are publication-bias adjusted. More trials and outcomes will be added over time.
More from AGI Musings
- Wiener's 1960 Alignment Warning Revisited: Patterning as a Control Problem — geoffreyirving · 2026-11-11
- Hinton Lectures 2026 kicks off with 'AI alignment: what's working and what's not' — DKokotajlo · 2026-11-11
- Why AI Slop Is So Frustrating to Read, per Zonder Labs — floguo · 2026-11-11
- Engineer: I used to write code all day, now I think about compute and AI economics — shakoistsLog · 2026-11-11
- "AI nerfs the based and nukes the mids": taste is the only moat left — dosco · 2026-11-11
- OpenAI board member Paul Christiano sees 'material risk' of AI takeover in 6-12 months — connoraxiotes · 2026-11-11