Nature Human Behaviour
How human–AI feedback loops alter human perceptual, emotional and social judgements
Key Takeaway
Across experiments with 1,401 participants, biased AI altered human perceptual, emotional, and social judgements and could amplify bias over time. Participants often underestimated the system influence.
AI-generated HIE assessmentStudy Design
multi experiment study
title and abstract
What the Study Found
Across experiments involving 1,401 participants, the authors find that interaction with biased AI can alter human perceptual, emotional and social judgements and amplify bias over time. Participants tended to underestimate the AI system's influence. The results span several tasks and algorithms, including a text-to-image system, but the persistence and broader generalizability of the effects require further study.
Original paper recordEvidence Strength
Evidence strengths
- 1,401 participants
- Multiple experiments and judgement domains
- Data and research design documentation available
Limitations
- Some tasks are laboratory abstractions
- Long-term persistence of the effect is not established
What This Does Not Prove
This record does not establish effects beyond the population, tasks, duration, and methods described by the paper. It should not be read as a universal causal claim.
HIE Interpretation
Direct experimental evidence that repeated interaction with biased AI can reshape and amplify human judgement biases.
AI-generated HIE assessment · AI assessment confidence 92Related Topics
- Human-AI Interaction
- Emotion
- Social Intelligence
- Cognitive Science
- Learning
Connected Claims
No public claims are connected to this paper yet.
Original Source
DOI 10.1038/s41562-024-02077-2 ↗