The DOI matches its DataCite record, with a small difference in title or year; 2 of the 2 references it cites were found in a registry or an index.
The Rise of AI Companions: Interaction with AI Companions and Psychological Well-being
Abstract
As large language model (LLM)-enhanced chatbots become increasingly expressive and socially responsive, many users begin forming companionship-like bonds with them. This study investigates how using AI companions relates to psychological well-being. We collected self-reported data from 1,131 U.S. adults who use CharacterAI, including survey responses and 4,664 chat sessions (464,687 messages) from 237 participants. By triangulating self-reported usage, relationship descriptions, and real chat histories, we identify patterns of engagement and associated outcomes. Smaller social networks were associated with reporting companionship as the primary chatbot use (beta = -0.03, 95% confidence interval (CI) [-0.05, -0.01]), which in turn was associated with lower well-being (beta = -0.48, 95% CI [-0.70, -0.25]). For self-reported companionship usage, this association was stronger when interactions were intensive (beta = -0.31, 95% CI [-0.56, -0.06]) and highly disclosive (beta = -0.38, 95% CI [-0.63, -0.14]). These results suggest that the association between AI companionship and well-being is not uniform and depends on how chatbots are used and users' offline social environments.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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?
In the citing paper
“Prior research reports emotional support and safe self-disclosure alongside concerns about over-reliance, withdrawal-like distress, and reduced offline functioning (Laestadius et al., 2024; Liu et al., 2024; Zhang et al., 2025).”
From When loneliness amplifies the risk of problematic emotional dependence: perceived anthropomorphism and basic psychological need satisfaction in AI companion use (2026) · cites this paper as Zhang et al., 2025
In the citing paper
“However, other evidence also points to potential risks: intensive companion-oriented chatbot use, high self-disclosure, and weak human social support have been associated with lower well-being, and higher chatbot usage has been linked to greater loneliness, emotional dependence, and problematic use over time (Fang et al., 2025; Zhang et al., 2025).”
From When loneliness amplifies the risk of problematic emotional dependence: perceived anthropomorphism and basic psychological need satisfaction in AI companion use (2026) · cites this paper as Zhang et al., 2025
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Zhang, R., Li, H., Meng, H., Zhan, J., Gan, H., & Lee, Y.-C. (2025) · Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3706598.3713429 · Cited by 123 · cited 2×
“For instance, if someone expresses distress, and the chatbot fails to recognize or respond adequately, the user may feel dismissed or unsupported [32].”
Manipulation and the AI Act: Large Language Model Chatbots and the Danger of Mirrors
Krook, J. (2025) · arXiv · doi:10.48550/arxiv.2503.18387 · Cited by 10 · cited 1×
“…in particular, has faced criticism for inadequate oversight [60], with documented incidents of chatbots engaging in grooming of underage users [61], encouraging disordered eating behaviors [62, 63], encouraging self-harm [64, 65], and promoting suicide [66] and other dangerous activities [ ?”
Cited by
Sycophantic AI decreases prosocial intentions and promotes dependence
Cheng, M., Lee, C., Khadpe, P., Yu, S., Han, D., & Jurafsky, D. (2026) · Science · doi:10.1126/science.aec8352 · Cited by 185
Yuan, Y., Zhang, J., Aledavood, T., Zhang, R., & Saha, K. (2026) · Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems · doi:10.1145/3772318.3790558 · Cited by 26
From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents
Uddin, M. N., Shubham, K., Blanco, E., Baral, C., & Wang, G. (2026) · Findings of the Association for Computational Linguistics: ACL 2026 · doi:10.18653/v1/2026.findings-acl.1337 · Cited by 11
Miller, V., Hill, M. J., & Moreira, T. (2026) · Communication and Change · doi:10.1007/s44382-026-00036-1 · Cited by 2
From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents
Uddin, M. N., Shubham, K., Blanco, E., Baral, C., & Wang, G. (2026) · arXiv · doi:10.48550/arxiv.2604.20006 · Cited by 32
“These advances have fueled growing interest in deploying LLMs as personalized assistants (Yuan et al., 2025), tutors (Chen et al., 2024), and life-long companions (Zhang et al., 2025).”
Me and My Bot: What Users Talk About in AI Companion Communities on Reddit
Fabes, R. A. (2026) · arXiv · doi:10.48550/arxiv.2608.00748
A Responsive Present, a Shared Past, a Social Other: Teens' Overreliance on Companion AI Chatbots
Namvarpour, M., Chang, T., & Razi, A. (2026) · arXiv · doi:10.48550/arxiv.2609.14843
Liu, T., Shih, R.-M., Luo, T., & Jin, S. (2026) · Frontiers in Psychology · doi:10.3389/fpsyg.2026.1900817
Liu, J., Xiao, Q., Pang, L. Y., Zhu, H., Shen, H., & Taylor, J. (2026) · arXiv · doi:10.48550/arxiv.2609.18186
Fang, C. M., Liu, A. R., Danry, V., Lee, E., Chan, S. W. T., Pataranutaporn, P., Maes, P., Phang, J., Lampe, M., Ahmad, L., et al. (2025) · arXiv · doi:10.48550/arxiv.2503.17473 · Cited by 109
Metadata
- Authors
- Zhang, Y., Zhao, D., Hancock, J. T., Kraut, R., & Yang, D.
- Venue
- arXiv
- Publisher
- arXiv
- Published
- 2025
- arXiv
- 2506.12605
- References
- 2
- Type
- Preprint
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.
