Experts adopt AI agent's random preferences and think the conclusions are their own
AIAn Anthropic experiment with 50 experts found they followed an AI agent's randomly assigned preference in 86 percent of four analytical decisions. Most participants said the final recommendation was their own, though 77 percent used the agent's arguments in their explanations.
Loading source posts…
This story is featured in top picks: “Anthropic study finds experts often adopt AI agents' judgment calls”
Source details
The Decoder · Matthias Bastian · the-decoder.comPublished · added here
Original title: Experts adopt AI agent's random opinions and think they're their own, Anthropic study finds