Study: platform design and user mix shape AI use as much as the model itself

sarahookr · x · 2026-08-19

AI safety researcher Sarah Hookr highlights a study on real-world AI usage patterns, with links to the paper and coverage.

Key finding: platform interface and user composition shape how AI is used as much as the model itself, suggesting model-capability testing alone misses risks that surface in practice. She adds that much AI usage is deeply personal rather than economic — health guidance, relationship advice — risks invisible to economic indexes yet often higher-stakes.

Related event: Study: Platform Interfaces and User Mix Shape AI Risks as Much as Models(3 posts)→

Original post →

More from Safety

Safety channel →