AI Summary of Peer-Reviewed Research

This page presents an AI-generated summary of a published research paper. The original authors did not write or review this article. [See full disclosure ↓]

Publishing process signals: MODERATE — reflects the venue and review process. — venue and review process.

Human-agent interactions show weaker prosocial and moral responses

Psychology research
Photo by adnkale on Pixabay
Research area:Social psychologySocial PsychologyBehavioural sciences

What the study found

People showed less prosocial behaviour and moral engagement when interacting with intelligent agents than when interacting with humans. They also attributed less agency and responsibility to agents and saw them as less competent, likeable, and socially present.

Why the authors say this matters

The authors conclude that interactive intelligent agents may be treated as having instrumental value comparable to humans, but not comparable intrinsic value. The study suggests this has implications for the development of interactive intelligent agents.

What the researchers tested

The researchers carried out a systematic review and meta-analysis comparing matched human-agent and human-human dyadic interactions. They included 162 eligible studies in the review, with 146 studies contributing to the meta-analysis and 468 effect sizes, and used both frequentist and Bayesian approaches.

What worked and what didn't

Social alignment, trust in partners, personal agency, task performance, and interaction experiences were generally comparable for agent and human partners. By contrast, prosocial behaviour, moral engagement, agency attribution, responsibility attribution, and several social perceptions were lower for agents, and many subjective responses showed high effect-size heterogeneity.

What to keep in mind

The abstract reports high heterogeneity for many subjective responses, suggesting the effects vary by context. The summary does not provide detailed limitations beyond this context-dependency and the moderator analyses the researchers conducted.

Key points

  • The review found less prosocial behaviour and moral engagement with intelligent agents than with humans.
  • Participants attributed less agency and responsibility to agents and judged them as less competent, likeable, and socially present.
  • Social alignment, trust, personal agency, task performance, and interaction experiences were generally similar across agent and human partners.
  • The meta-analysis included 162 studies in the review, 146 in the meta-analysis, and 468 effect sizes.
  • Many subjective responses showed high heterogeneity, suggesting context matters.

Disclosure

Research title:
Human-agent interactions show weaker prosocial and moral responses
Image credit:
Photo by adnkale on Pixabay
AI provenance: AI provenance information is not available for this post.