CoChat
Sign Up
Cited by 39Social Robot Interaction and HRIAI in Service InteractionsAttachment and Relationship Dynamics

Love, marriage, pregnancy: Commitment processes in romantic relationships with AI chatbots

Open access
Djufril, Ray, Frampton, Jessica R., Knobloch-Westerwick, Silvia·2025·Computers in Human Behavior: Artificial Humans
DOI 10.1016/j.chbah.2025.100155
Open this paper in CoChat — free account

Abstract

An inductive thematic analysis examined written responses from 29 individuals using the romantic relationship function of the social chatbot Replika. Findings indicate that most of these users feel an emotional connection to the bot, that the bot meets their needs when there are no technical issues, and that interactions with the bot are often different from (and sometimes better than) interactions with humans. All these factors impact users’ commitment to their human-chatbot relationship. Additionally, the study explored how users navigated a time of relational transition, specifically a period of erotic roleplaying censorship. Participants experienced intense emotional responses, but many were guarded from negativity bias toward their AI partner because of the ability to blame developers. These findings are discussed in light of the investment model, the computers are social actors paradigm, social affordances, and relational turbulence theory. • A bot's romantic relationship function can induce intense emotional connection. • Bot romance is often perceived as different from, or better than, human romance. • Increased commitment to bots involves high emotional investment and bonding rituals. • Users navigate relational turbulence in bot romances by blaming developers.

Integrity signals

Registry

Crossref · 10.1016/j.chbah.2025.100155

Cites

49 of 49 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Claims the paper makes

We couldn’t retrieve this paper’s full text, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Roleplay environments such as Character.AI do not simply extend traditional fan cultures; they represent a fundamental shift in how relationships, identities, and emotional attachments are technically mediated.

    “Roleplay environments such as Character.AI do not simply extend traditional fan cultures; they represent a fundamental shift in how relationships, identities, and emotional attachments are technically mediated”

    From Toward an Ethic of Synthetic Relationality: Identity, Intimacy, and Risk in AI-Mediated Roleplay Environments (2025) · cites this paper as Djufril, Frampton, and Knobloch-Westerwick 2025

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Djufril, R., Frampton, J. R., & Knobloch-Westerwick, S.
Affiliations
  • Technische Universität Berlin (DE)
  • University of Tennessee at Knoxville (US)
Corresponding author
Frampton, Jessica R.
ORCID
Frampton, Jessica R.
Venue
Computers in Human Behavior: Artificial Humans
Publisher
Elsevier BV
Published
2025-05
Volume
4
Pages
100155
ISSN
2949-8821 (ISSN-L)
DOI
10.1016/j.chbah.2025.100155
OpenAlex
W4409451912
Cited by
33
References
49
Citations by year
2026: 27 · 2025: 6
FWCI
36.56
Type
journal article
Language
en
Open access
Licence
Locations
  • Computers in Human Behavior Artificial Humans (publishedVersion, cc-by-nc-nd)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Social Sciences › Psychology › Social Psychology › Social Robot Interaction and HRI
  • Physical Sciences › Computer Science › Artificial Intelligence › AI in Service Interactions
  • Social Sciences › Psychology › Social Psychology › Attachment and Relationship Dynamics
Keywords
AI chatbots, relationship commitment, Replika, relational turbulence, investment model, Computers Are Social Actors, social affordances, blame attribution

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review