AI-generated vs. human therapist feedback in internet-based CBT for social anxiety disorder: A six-arm randomized controlled trial with deception design
5 affiliations
- 1Department of Psychology, Stockholm University, Stockholm, Sweden
- 2Department of Behavioural Sciences and Learning, Linköping University, Linköping, Sweden
- 3Department of Health, Education and Technology, Luleå University of Technology, Luleå, Sweden
- 4Lusófona University, Digital Human-Environment Interaction Labs, Lisboa, Portugal
- 5University College, Korea University, Seoul, 02481, South Korea
Abstract
Introduction: Guided internet-delivered cognitive behavioral therapy (ICBT) is effective for social anxiety disorder (SAD), but therapist-delivered feedback limits its scalability. Whether AI-generated feedback can match human feedback, and whether patients' expectations about the source shape outcomes is unknown.
Method: In a six-arm randomized controlled trial, 263 adults with SAD received psychologist-reviewed AI-generated or human feedback. They were told it came from a human or an AI which was correct or incorrect information (four fully crossed feedback arms), or were allocated to unguided self-help or a waitlist. The Social Phobia Inventory (SPIN) was analyzed using linear mixed models, complemented by Bayesian estimation. Secondary outcomes included therapeutic alliance, treatment credibility, negative effects, depression, and anxiety.
Results: Active treatment yielded large reductions in social anxiety relative to waitlist (d = −1.23). Before treatment, participants rated it as more credible when they expected feedback from a human rather than AI. AI and human feedback did not however statistically differ in symptom reduction (d = 0.10, 95% CI [−3.62, 5.91]); non-inferiority was not established. Neither the actual source, participants' beliefs, nor their congruence influenced symptoms or alliance. Depression, anxiety, and negative effects did not differ by source. Stronger alliance predicted subsequent symptom reduction, but not the reverse.
Conclusions: Psychologist-reviewed AI-generated feedback yielded reductions in social anxiety symptoms that did not differ significantly from human feedback, without greater harm, and despite participants' initial expectation that AI would be less credible. The barrier to AI-supported ICBT may therefore lie in expectations rather than outcomes.
Trial registration: ISRCTN11343758.
Related
- The Dose-Response Investigation of Docosahexaenoic acid (DRI-DHA) Trial: A randomized controlled trial demonstrates DHA feedback inhibition of its own synthesis
- Efficacy of real-time feedback exercise therapy in patients following total hip arthroplasty: a pilot cluster-randomized controlled trial
- Multi-Layered Recursive Cyber Deception Architecture Powered by True Cosmic-Ray Inter-Particle Randomized Shuffling (Particle-on-Particle) over Decentralized mDOM Meshes
- Comparative efficacy of brain-computer interface training with different feedback modalities for post-stroke upper-limb motor recovery: a systematic review and network meta-analysis of randomized controlled trials
- Exploring Artificial Intelligence-Generated Feedback for Pharmacy Students †
BibTeX
@article{hlynsson2026aigenerated,
title = {{AI-generated vs. human therapist feedback in internet-based CBT for social anxiety disorder: A six-arm randomized controlled trial with deception design}},
author = {Jón Ingi Hlynsson and Gerhard Andersson and Per Carlbring},
journal = {Internet Interventions},
year = {2026},
month = oct,
doi = {10.1016/j.invent.2026.101012},
url = {https://doi.org/10.1016/j.invent.2026.101012},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - AI-generated vs. human therapist feedback in internet-based CBT for social anxiety disorder: A six-arm randomized controlled trial with deception design
AU - Jón Ingi Hlynsson
AU - Gerhard Andersson
AU - Per Carlbring
PY - 2026
DA - 2026/10/16
JO - Internet Interventions
DO - 10.1016/j.invent.2026.101012
UR - https://doi.org/10.1016/j.invent.2026.101012
AB - Introduction: Guided internet-delivered cognitive behavioral therapy (ICBT) is effective for social anxiety disorder (SAD), but therapist-delivered feedback limits its scalability. Whether AI-generated feedback can match human feedback, and whether patients' expectations about the source shape outcomes is unknown. Method: In a six-arm randomized controlled trial, 263 adults with SAD received psychologist-reviewed AI-generated or human feedback. They were told it came from a human or an AI which was correct or incorrect information (four fully crossed feedback arms), or were allocated to unguided self-help or a waitlist. The Social Phobia Inventory (SPIN) was analyzed using linear mixed models, complemented by Bayesian estimation. Secondary outcomes included therapeutic alliance, treatment credibility, negative effects, depression, and anxiety. Results: Active treatment yielded large reductions in social anxiety relative to waitlist (d = −1.23). Before treatment, participants rated it as more credible when they expected feedback from a human rather than AI. AI and human feedback did not however statistically differ in symptom reduction (d = 0.10, 95% CI [−3.62, 5.91]); non-inferiority was not established. Neither the actual source, participants' beliefs, nor their congruence influenced symptoms or alliance. Depression, anxiety, and negative effects did not differ by source. Stronger alliance predicted subsequent symptom reduction, but not the reverse. Conclusions: Psychologist-reviewed AI-generated feedback yielded reductions in social anxiety symptoms that did not differ significantly from human feedback, without greater harm, and despite participants' initial expectation that AI would be less credible. The barrier to AI-supported ICBT may therefore lie in expectations rather than outcomes. Trial registration: ISRCTN11343758.
SN - 2214-7829
ER -