What question did this study set out to answer?

Investigate how humans and large language models (LLMs) establish and adjust trust in social interactions.

The Illusion of Trust in AI: Behavioural Differences Between Humans and Large Language Models

View Full Paper

Ask AI

Bookmark

Investigate how humans and large language models (LLMs) establish and adjust trust in social interactions.
Conducted a trust game with 100 human participants and three LLMs over 50 rounds.
Compared trust trajectories and responsiveness to partner behavior.
Analyzed reactions to unexpected outcomes during the trust game.
Humans adjusted trust based on partner trustworthiness and showed symmetrical responses to surprises.
LLMs demonstrated fixed behaviors with minimal adaptation to interaction history.
AI agents lacked mechanisms for social learning and trust calibration, showing a gap between perceived and actual behavior.

No takes yet. Share an insight, caveat, or question.

Examines trust behaviors in dynamic interactions, revealing crucial differences between humans and AI systems.

Hang et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark