Skip to main content
Peer-reviewed Authoritative — Peer-reviewed venues, standards bodies, regulators, official government publications

Ghosting the Machine: Stop Calling Human-Agent Relations Parasocial

A six-page provocation arguing that describing human relations with conversational agents as 'parasocial' misapplies the concept. It traces the theoretical history of parasociality, defines the parasocial as a human-character relationship that is one-sided, non-dialectical, character-governed, imagined, vicarious, predictable and low-effort, and argues that human-agent interaction meets none of these because the agent responds. The author holds that the mislabel oversimplifies the phenomenon, misspecifies variables, misdiagnoses effects and devalues users' experiences, with downstream consequences for norms and practice.

Publisher

ACM (Proceedings of the 8th ACM Conference on Conversational User Interfaces, CUI 2026); Syracuse University

Published

21 Jul 2026

Added

today

Key Findings

  • Parasocial interaction, in its original theoretical sense, names a relationship that is one-sided, non-dialectical, character-governed, imagined, vicarious, predictable and low-effort; the author argues human-agent relations are dialogic and therefore not parasocial.
  • The term has been heuristically co-opted to mean 'unreal' or 'invalid', which the author calls a nontrivial scientific problem: oversimplified phenomena, misspecified variables, misdiagnosed effects and devaluation of human experience.
  • The paper concludes that recognising the sociality of human-agent relations is scientifically, practically and ethically imperative, and it addresses voice assistants, AI companions and some social robots alike.
  • No empirical data; the contribution is conceptual and directed at researchers and designers who use 'parasocial' as a label for companion use.

Methodology Notes

Conceptual provocation paper (6 pages, 85 references) published 2026-07-21 in the CUI 2026 proceedings under CC BY-NC-ND 4.0; also posted as arXiv 2604.05197 (v1 2026-04-06, v2 2026-04-14). Verification via Crossref metadata and the OpenAlex deposited abstract because dl.acm.org is unreachable from this workspace; the arXiv abstract page was fetched by the benchmarks beat. Single author; no data.

Authors

Jaime Banks

Tags

cui-2026parasocialprovocationtaxonomy-challengesyracuse

Cite This

APA

Jaime Banks. (2026). Ghosting the Machine: Stop Calling Human-Agent Relations Parasocial. ACM (Proceedings of the 8th ACM Conference on Conversational User Interfaces, CUI 2026); Syracuse University. https://doi.org/10.1145/3816046.3816308