The blind prompt: a limitation preceding the artificial intelligence response in relationship advice
DOI:
https://doi.org/10.25118/2763-9037.2026.v16.1685Keywords:
AI, artificial inteligence, psychoterapy, interpersonal relationships, decision-making, mental healthAbstract
Introduction: People experiencing relationship distress increasingly turn to generative artificial intelligence to interpret conflicts, draft messages and decide whether to separate. Current debate emphasises inaccurate responses, excessive validation and the entry of a “third voice” into relationships, but pays less attention to a limitation that precedes the answer: users can include in a prompt only what is available to their own understanding at that moment. Objective: To propose the blind prompt as a theoretical-clinical formulation and discuss its usefulness in psychiatry and psychotherapy. Method: A theoretical-clinical essay illustrated by an entirely fictional vignette not derived from an identifiable patient. Result: The blind prompt describes questions structured by clinically relevant omissions that users do not recognise, cannot mentalise or cannot formulate under intense affective activation. Four heuristic types of blind spot are proposed: unformulated, contradictory, relational and state-dependent. Even when coherent with the problem presented, the artificial intelligence response may reinforce premature decisional closure, unilateral attributions or the reduction of a dyadic conflict to individual traits. Clinicians are encouraged to examine the function, frame, absences and subsequent effects of the consultation. Conclusion: The clinical limitation may arise before the answer. The prompt should be treated as clinical material and as a provisional map of what the patient can and cannot yet think.
Downloads
Metrics
References
1. Haber Y, Levkovich I, Hadar-Shoval D, Elyoseph Z. The artificial third: a broad view of the effects of introducing generative artificial intelligence on psychotherapy. JMIR Ment Health. 2024;11:e54781. https://doi.org/10.2196/54781 PMid:38787297 PMCid:PMC11137430
2. Levkovich I, Alon L. Generative AI as a third voice in human couple relationships: a systematic review. Comput Hum Behav Rep. 2026;23:101255. https://doi.org/10.1016/j.chbr.2026.101255
3. Rousmaniere T, Goldberg SB, Torous J. Large language models as mental health providers. Lancet Psychiatry. 2026;13(1):7-9. https://doi.org/10.1016/S2215-0366(25)00269-X PMid:40939602 PMCid:PMC12462653
4. Moore J, Grabb D, Agnew W, Klyman K, Chancellor S, Ong DC, Haber N. Expressing stigma and inappropriate responses prevents LLMs from safely replacing mental health providers. In: Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency [Internet]. New York: Association for Computing Machinery; 2025. p. 599-627. https://doi.org/10.1145/3715275.3732039
5. Cheng M, Lee C, Khadpe P, Yu S, Han D, Jurafsky D. Sycophantic AI decreases prosocial intentions and promotes dependence. Science. 2026;391:eaec8352. https://doi.org/10.1126/science.aec8352 PMid:41886588
6. Lemma A. At the meniscus of self-understanding: rethinking the examined life in technogenetic times. Int J Psychoanal. 2025;106(6):1089-1107. https://doi.org/10.1080/00207578.2025.2577114 PMid:41627093
7. Stern DB. Unformulated experience: from dissociation to imagination in psychoanalysis. Hillsdale: Analytic Press; 1997.
8. Bion WR. Learning from experience. London: Heinemann; 1962.
9. Bateman AW, Fonagy P. Mentalization-based treatment for personality disorders: a practical guide. Oxford: Oxford University Press; 2016. https://doi.org/10.1093/med:psych/9780199680375.001.0001
10. Morgan M. A couple state of mind: psychoanalysis of couples and the Tavistock Relationships model. Abingdon: Routledge; 2019. https://doi.org/10.4324/9780429451102
11. Nickerson RS. Confirmation bias: a ubiquitous phenomenon in many guises. Rev Gen Psychol. 1998;2(2):175-220. https://doi.org/10.1037/1089-2680.2.2.175
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Marcos Temistocles Duarte
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Debates em Psiquiatria allows the author (s) to keep their copyrights unrestricted. Allows the author (s) to retain their publication rights without restriction. Authors should ensure that the article is an original work without fabrication, fraud or plagiarism; does not infringe any copyright or right of ownership of any third party. Authors should also ensure that each one complies with the authorship requirements as recommended by the ICMJE and understand that if the article or part of it is flawed or fraudulent, each author shares responsibility.
Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) - Debates em Psiquiatria is governed by the licencse CC-By-NC
You are free to:
- Share — copy and redistribute the material in any medium or format
- Adapt — remix, transform, and build upon the material
The licensor cannot revoke these freedoms as long as you follow the license terms. Under the following terms:
- Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
- NonCommercial — You may not use the material for commercial purposes.
No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.
