Exploring User Security and Privacy Attitudes and Concerns Toward the Use of General-Purpose LLM Chatbots for Mental Health

Jabari Kwesi; Jiaxun Cao; Pardis Emami-Naeini; Riya Manchanda

REVIEW 3 major objections 4 minor 126 references

Exploring User Security and Privacy Attitudes and Concerns Toward the Use of General-Purpose LLM Chatbots for Mental Health

T0 review · 3 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Users grant mental-health chatbots the legal trust of therapists.

desk verdict Useful qualitative snapshot with a new concept, but the abstract's empathy-to-accountability claim runs ahead of the data. read the letter →

arxiv 2507.10695 v1 pith:6RARBFOU submitted 2025-07-14 cs.CY cs.AIcs.CRcs.ETcs.HC

Jabari Kwesi , Jiaxun Cao , Riya Manchanda , Pardis Emami-Naeini This is my paper

classification cs.CYcs.AIcs.CRcs.ETcs.HC

keywords LLMchatbotsmentalhealthprivacysecurityattitudesHIPAAmisconceptionintangiblevulnerabilitysemi-structuredinterviewsexpectationschatbotregulation

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that people who use general-purpose LLM chatbots for mental-health support hold systematically mistaken beliefs about the privacy and security of those conversations. In 21 semi-structured interviews with U.S. adults, users attributed the chatbots' human-like empathy to human-like accountability, and some believed their disclosures were protected by health-privacy regulations such as HIPAA even though most of these tools are not covered entities. The authors also introduce 'intangible vulnerability': users rate emotional disclosures as deeply personal yet less at risk than financial or location data, because they cannot picture a concrete way for such information to be exploited. The result, if correct, is that a meaningful share of the most sensitive disclosures are made under false legal assumptions, and safer design has to address the gap between perceived and actual protection rather than treating privacy as an individual user's responsibility.

What carries the argument

The machinery of the paper is the interview study itself, and the conceptual instrument it builds is the term 'intangible vulnerability'—the finding that users acknowledge emotional disclosures to be deeply personal but undervalue them relative to tangible information such as credit-card numbers or home addresses, because they cannot imagine a concrete exploitation path. The concept does the argumentative work of explaining why users who are otherwise privacy-conscious will strip their names from a prompt yet still disclose trauma in rich detail, and why roughly half of the sample adopted no protective measures at all. It also anchors the paper's design recommendations: just-in-time warnings when a conversation appears to be a mental-health disclosure, ephemeral storage as a default rather than an opt-in, and targeted third-party audits of data handling for tools likely to collect health-like information.

What would settle it

A concrete check would be a controlled vignette or logged-usage study: let consenting users interact with an empathetically worded chatbot in one condition and a neutral but functionally identical chatbot in another, then ask both groups whether their chats are legally protected like therapist conversations; if the neutral group holds the same HIPAA beliefs as the empathetic group, the empathy-to-accountability mechanism is contradicted. A complementary comparison would take a consenting panel's real chat logs and test whether people who say they withhold identifiers actually do so at the prompt level, and whether their stated beliefs about data handling match the platform's actual retention and sharing settings.

Watch

Extended reading notes

Core claim

Central finding: privacy attitudes toward general-purpose LLM chatbots used for mental health are driven by two misperceptions that the paper labels separately. The first is a conflation of empathy with accountability: participants described responses as 'nonjudgmental' and 'personalized,' compared the tool favorably with therapists, and 7 of 21 assumed that their conversations were governed by regulations like HIPAA or by doctor-patient confidentiality. The second misperception is a sensitivity hierarchy in which emotional disclosures sit below financial or locational data: interviewees worried about credit-card numbers, addresses, employers, or insurers, but treated anxiety triggers, trauma, or body-image struggles as comparatively hard to exploit. Together these misperceptions produce what the authors call 'intangible vulnerability'—emotional or psychological disclosures are considered most private yet least protected, because users cannot map them onto familiar harm scenarios such as identity theft or doxxing. The paper further reports that roughly half of participants tried some protective practice, mostly by omitting names and other identifiers, while half did not, citing trust, resignation, or altruism.

Load-bearing premise

The load-bearing premise is that participants' self-reported accounts of their chatbot use, privacy beliefs, and protective behaviors match what they actually think and do; the paper itself notes that no direct observation or usage logs were collected to corroborate those reports.

Editorial extensions

If this is right

If the HIPAA misperception is widespread, then a large share of mental-health disclosures are made under a false legal assumption, since most general-purpose LLM chatbots are not covered entities under that law.
Users who believe their chats are legally protected have little reason to seek out protective settings, which is consistent with the paper's observation that most participants never read the privacy policy and half did not use any mitigation.
Because emotional disclosures feel less exploitable than financial or location data, data-minimization strategies such as removing names and timelines are an unreliable safeguard; users will still expose the information that makes them identifiable in context.
The paper's recommendations follow from the mechanism: default ephemeral storage, automatic prompts reminding users that the tool is not a licensed therapist, and audited data handling would protect users where self-directed vigilance has been shown to fail.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

An extension the paper does not test: if the empathy-to-accountability conflation is causal, then increasing personification of a chatbot—through voice, memory, an avatar, or a backstory—should increase users' belief in legal protection; this could be measured experimentally by varying only the personification cues.
The intangible-vulnerability mechanism may generalize beyond mental health to other high-intimacy disclosures such as relationship troubles, sexuality, or political beliefs, where no concrete monetization path is visible; a vignette study comparing willingness to share emotional versus financial information would test this.
Because the study is U.S.-specific, the authors leave open how the same conflation behaves under privacy regimes with broader data-protection rules; a cross-country replication could show whether false HIPAA confidence is replaced by a different but equally misplaced belief when a general data-protection law exists.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Plainly: this is a modest but genuinely useful qualitative study. Its main empirical contribution is the documentation of mistaken HIPAA beliefs among a notable minority of users (7/21) of general-purpose LLM chatbots for mental health, plus the introduction of 'intangible vulnerability' as a label for why people treat emotional disclosures as lower-risk than financial or location data. That concept is worth taking seriously, and the recommendations for harm reduction (just-in-time warnings, ephemeral storage by default, third-party audits) follow sensibly from the interviews.

The paper is well run. The methods are transparent, the authors avoided priming with the word 'privacy' in recruitment, they report thematic saturation, and they make the interview guide and codebook available. For a qualitative study with 21 participants, that is solid practice. I believe the finding about HIPAA misconceptions, in particular, is a real contribution to the privacy and HCI literature.

Where the paper goes wrong is in the abstract and the discussion. The claim that participants 'conflated the human-like empathy exhibited by LLMs with human-like accountability' is not supported by the evidence presented. Look at the quotes: P21 assumed a clickwrap agreement gave the same protections as a therapist; P7 cited the chatbot's 'database of research'; P13 argued that sensitive health information should be under the same legal umbrella. These are mistaken beliefs about legal scope, not evidence that empathy caused them. No participant is quoted making the empathy-to-accountability connection. The robust finding is narrower: some users falsely believe HIPAA covers these chats. The conflation is an interpretive leap by the authors, and it is load-bearing in the abstract because it motivates the regulatory recommendations. The paper still supports harm-reduction recommendations without it, but the abstract oversells the mechanism.

Also, the 'hypothetical agreement of 100%' for coding reliability is not a meaningful metric. The authors should either do a proper inter-coder reliability check or simply report how disagreements were resolved. The decision not to share raw transcripts is understandable for privacy, but it limits verification; that has to be acknowledged as a trade-off, not a virtue.

All that said, these are fixable problems. The study's central empirical finding is plausible and useful, and the authors are clear about the limitations of self-report and the Prolific sample. This is not a breakthrough, but it deserves to be read by people working on mental health AI and privacy. If I were an editor, I would send it out to review, with a request to revise the abstract and the wording of the coding reliability paragraph.

Referee Report

3 major / 4 minor

Summary. This paper reports a qualitative interview study with 21 U.S. adults recruited via Prolific who use general-purpose LLM chatbots (e.g., ChatGPT, Gemini, Replika) for mental health support. Through inductive thematic analysis, the authors describe participants' motivations (cost, accessibility, perceived neutrality), limited awareness of data handling, mistaken beliefs that HIPAA or therapist-style confidentiality applies, mitigation practices such as de-identification and VPN use, and expectations about responsibility and regulation. The paper introduces the concept of 'intangible vulnerability' and closes with harm-reduction recommendations (contextual nudges, ephemeral storage, targeted audits).

Significance. If the central empirical findings are taken at face value, the paper addresses an important and understudied area: users may disclose sensitive mental-health information to general-purpose LLM chatbots while holding false assumptions about legal protection. Strengths include an inductive coding process with reported saturation, an explicit ethical protocol, and the release of study materials and a codebook. The proposed concept of 'intangible vulnerability' is a useful interpretive lens. However, the abstract's headline mechanism—that empathy was conflated with accountability—is not directly supported by the quoted evidence, and the reliability reporting is unclear; the supported core is narrower but still meaningful.

major comments (3)

[Abstract; §5.3] The abstract and Section 5.3 claim that participants 'conflated the human-like empathy exhibited by LLMs with human-like accountability' and therefore believed HIPAA protected their chats. The quoted evidence does not establish this causal/interpretive link: P7 invokes a 'database of research,' P21 assumes clickwrap agreements, P13 applies a general 'same umbrella' argument about sensitive health information, and P16/P15 emphasize non-judgmental or neutral interaction, not legal accountability. The data support a narrower, still important claim that 7/21 participants held false beliefs about HIPAA coverage and that some participants expected therapy-like confidentiality. Please revise the abstract and discussion to state the narrower claim, or provide direct participant statements (or a clearly labeled analytic inference) showing that empathy perception drove accountability beliefs.
[§3.2 Qualitative Analysis] The reliability statement 'leading to a hypothetical agreement of 100%' is not a meaningful coding-consistency measure. If all disagreements were resolved by discussion, there is no independent agreement metric; 'hypothetical' makes the claim unverifiable. This matters because the MC7 theme (HIPAA misconception, 7/21) is load-bearing for the abstract's central claim. Please report the coding process transparently (e.g., number of transcripts double-coded, initial agreement, how disagreements were resolved, and any audit trail) or explicitly state that no quantitative inter-rater reliability was computed.
[§3.3 Limitations] The paper's conclusions sometimes generalize beyond the sample and method. The authors acknowledge self-selection via Prolific, self-reported usage, and lack of observational logs, yet the abstract and conclusion phrase findings as 'participants conflated...' and 'demonstrated that...' without hedging. Since the central empirical premise is self-reported attitudes, the conclusions should consistently be framed as perceptions reported in interviews, with the limitations restated in the abstract or conclusion.

minor comments (4)

[§4.1] There is a grammatical error in the passage about P12: 'Emotional disclosures seemed less exploitable to he' should read 'to him.'
[§1] The sentence-initial 'Moreso' in the Introduction is nonstandard; consider 'Moreover'.
[Table 1] The column headed 'Used AI Chatbot' lists tools such as ChatGPT, Replika, and Grok; consider labeling it 'Used general-purpose LLM chatbot' to match the eligibility criterion and the paper's terminology.
[§5.5] The phrase 'a particularly tragic such case' is awkward; consider rephrasing to 'a particularly tragic case.'

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; the interview study is inductive and its conclusions are not equivalent to its inputs.

full rationale

This paper reports a qualitative interview study with 21 participants and makes no mathematical or predictive claims, fits no parameters, and derives no quantities from its own outputs. The central concepts, such as "intangible vulnerability" (Section 5.2), are interpretive labels applied after data collection to patterns in participant statements, not assumptions imported into the analysis; the paper explicitly describes an inductive thematic analysis in which themes "emerge directly from the data rather than applying predefined frameworks" (Section 3.2). No load-bearing self-citations or imported uniqueness theorems appear; references to prior work (e.g., Tufekci, Brandimarte et al., Acquisti and Grossklags) are used as points of comparison, not as forced premises. The skeptical concern that the "empathy-to-accountability" conflation is asserted rather than directly evidenced is a correctness or evidentiary criticism, not a circularity: the abstract's interpretive claim is not equivalent by construction to any participant quote or to the interview protocol. Likewise, the acknowledged limitations (self-report, recall bias, social desirability, no usage logs, unavailable transcripts) weaken evidentiary strength but do not create a loop in which a conclusion is defined into existence by its inputs. The paper is self-contained against external benchmarks in the sense appropriate to thematic analysis: its findings are grounded in reported data, its limitations are disclosed, and no claim reduces to a prior commitment of the authors.

Assumptions & free parameters 0 free parameters · 4 assumptions · 1 invented entities

The study is qualitative and contains no fitted models or free parameters. Its central claims rest on self-reported interview data, interpretive coding, and an assumed saturation threshold. The only introduced concept, 'intangible vulnerability,' is an interpretive label without independent verification.

assumptions (4)

domain assumption Participants' self-reported accounts during interviews accurately reflect their real-world privacy attitudes and behaviors.
All central findings rely on self-report without observational usage logs; acknowledged as a limitation in Section 3.3.
domain assumption Thematic saturation reached after 17 interviews is sufficient to capture the range of relevant attitudes among the target population.
The authors assert saturation without a formal saturation analysis (Section 3), so the completeness of the thematic map is assumed.
domain assumption The coding process produced a valid and consistent representation of the interview data.
Inter-rater reliability is reported only as a 'hypothetical agreement of 100%' after discussion (Section 3.2), which does not constitute a standard reliability metric.
domain assumption The Prolific-recruited sample provides meaningful diversity and sufficient representation for the study's claims.
Participants are self-selected and likely tech-savvy; the authors note the sample is not representative in Section 3.3.

invented entities (1)

Intangible vulnerability
purpose: Conceptual label for the phenomenon where emotional or psychological disclosures are undervalued relative to tangible data such as financial or location information, reducing perceived privacy risk.
Introduced in Section 5.2 as a novel theme from the interviews; no independent falsifiable evidence is provided, making it an interpretive construct rather than a measured quantity.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Exploring User Security and Privacy Attitudes and Concerns Toward the Use of General-Purpose LLM Chatbots for Mental Health." pith.science (2026). https://pith.science/paper/6RARBFOU

@misc{pith2026250710695,
  author       = {Pith},
  title        = {Pith review of: Exploring User Security and Privacy Attitudes and Concerns Toward the Use of General-Purpose LLM Chatbots for Mental Health},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6RARBFOU}},
  note         = {Machine review of arXiv:2507.10695}
}

read the original abstract

Individuals are increasingly relying on large language model (LLM)-enabled conversational agents for emotional support. While prior research has examined privacy and security issues in chatbots specifically designed for mental health purposes, these chatbots are overwhelmingly "rule-based" offerings that do not leverage generative AI. Little empirical research currently measures users' privacy and security concerns, attitudes, and expectations when using general-purpose LLM-enabled chatbots to manage and improve mental health. Through 21 semi-structured interviews with U.S. participants, we identified critical misconceptions and a general lack of risk awareness. Participants conflated the human-like empathy exhibited by LLMs with human-like accountability and mistakenly believed that their interactions with these chatbots were safeguarded by the same regulations (e.g., HIPAA) as disclosures with a licensed therapist. We introduce the concept of "intangible vulnerability," where emotional or psychological disclosures are undervalued compared to more tangible forms of information (e.g., financial or location-based data). To address this, we propose recommendations to safeguard user mental health disclosures with general-purpose LLM-enabled chatbots more effectively.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

126 extracted references · 74 canonical work pages

[1]

An overview of the features of chatbots in mental health: A scoping review

Alaa A Abd-Alrazaq, Mohannad Alajlani, Ali Abdallah Alalwan, Brid- gette M Bewick, Peter Gardner, and Mowafa Househ. An overview of the features of chatbots in mental health: A scoping review. Inter- national journal of medical informatics, 132:103978, 2019

2019
[2]

Perceptions and opinions of patients about mental health chatbots: scoping review

Alaa A Abd-Alrazaq, Mohannad Alajlani, Nashva Ali, Kerstin De- necke, Bridgette M Bewick, and Mowafa Househ. Perceptions and opinions of patients about mental health chatbots: scoping review. Journal of medical Internet research, 23(1):e17828, 2021

2021
[3]

Large language models associate muslims with violence.Nature Machine Intelligence, 3:461–463, 06 2021

Abubakar Abid, Maheen Farooqi, and James Zou. Large language models associate muslims with violence.Nature Machine Intelligence, 3:461–463, 06 2021

2021
[4]

Imagined communities: Aware- ness, information sharing, and privacy on the facebook

Alessandro Acquisti and Ralph Gross. Imagined communities: Aware- ness, information sharing, and privacy on the facebook. In Interna- tional workshop on privacy enhancing technologies , pages 36–58. Springer, 2006

2006
[5]

Privacy attitudes and pri- vacy behavior: Losses, gains, and hyperbolic discounting

Alessandro Acquisti and Jens Grossklags. Privacy attitudes and pri- vacy behavior: Losses, gains, and hyperbolic discounting. In Eco- nomics of information security, pages 165–178. Springer, 2004

2004
[6]

Privacy and rationality in individual decision making

Alessandro Acquisti and Jens Grossklags. Privacy and rationality in individual decision making. IEEE security & privacy , 3(1):26–33, 2005

2005
[7]

What can behavioral eco- nomics teach us about privacy? In Digital privacy, pages 363–378

Alessandro Acquisti and Jens Grossklags. What can behavioral eco- nomics teach us about privacy? In Digital privacy, pages 363–378. Auerbach Publications, 2007

2007
[8]

Assessing the effectiveness of chatgpt in delivering mental health support: a qualitative study.Journal of Multidisciplinary Healthcare, pages 461–471, 2024

Fahad Alanezi. Assessing the effectiveness of chatgpt in delivering mental health support: a qualitative study.Journal of Multidisciplinary Healthcare, pages 461–471, 2024

2024

Show all 126 references

[9]

New public library technology survey report details digital equity roles, 2024

American Library Association. New public library technology survey report details digital equity roles, 2024

2024
[10]

Media literacy education for adult audiences: Demystifying ai

American Library Association. Media literacy education for adult audiences: Demystifying ai. Video, 2024

2024
[11]

C., & tucker, ce the digital privacy paradox: Small money, small costs, small talk, 2018

S Catalini Athey. C., & tucker, ce the digital privacy paradox: Small money, small costs, small talk, 2018

2018
[12]

Evaluating artifi- cial intelligence responses to public health questions

John W Ayers, Zechariah Zhu, Adam Poliak, Eric C Leas, Mark Dredze, Michael Hogarth, and Davey M Smith. Evaluating artifi- cial intelligence responses to public health questions. JAMA network open, 6(6):e2317517–e2317517, 2023

2023
[13]

Can an AI Chatbot be your therapist? https://business.yougov.com/content/49480-can-an-ai-chatbot- be-your-therapist

Bhavika Bansal. Can an AI Chatbot be your therapist? https://business.yougov.com/content/49480-can-an-ai-chatbot- be-your-therapist
[14]

Putting the privacy paradox to the test: Online privacy and security behaviors among users with technical knowledge, privacy awareness, and financial resources

Susanne Barth, Menno DT de Jong, Marianne Junger, Pieter H Har- tel, and Janina C Roppelt. Putting the privacy paradox to the test: Online privacy and security behaviors among users with technical knowledge, privacy awareness, and financial resources. Telematics and informatic...

2019
[15]

Psychiatrists’ experiences and opinions of generative artificial intelligence in mental healthcare: An online mixed methods survey

Charlotte Blease, Abigail Worthen, and John Torous. Psychiatrists’ experiences and opinions of generative artificial intelligence in mental healthcare: An online mixed methods survey. Psychiatry Research, 333:115724, 2024

2024
[16]

Misplaced confidences: Privacy and the control paradox

Laura Brandimarte, Alessandro Acquisti, and George Loewenstein. Misplaced confidences: Privacy and the control paradox. Social psy- chological and personality science, 4(3):340–347, 2013

2013
[17]

Caring is not enough: The importance of internet skills for online privacy protection

Moritz Büchi, Natascha Just, and Michael Latzer. Caring is not enough: The importance of internet skills for online privacy protection. Information, Communication & Society, 20(8):1261–1278, 2017

2017
[18]

Pilot randomised con- trolled trial of help4mood, an embodied virtual agent-based system to support treatment of depression

Christopher Burton, Aurora Szentagotai Tatar, Brian McKinstry, Colin Matheson, Silviu Matu, Ramona Moldovan, Michele Macnab, Elaine Farrow, Daniel David, Claudia Pagliari, et al. Pilot randomised con- trolled trial of help4mood, an embodied virtual agent-based system to suppor...

2016
[19]

Assessing the usability of a chatbot for mental health care

Gillian Cameron, David Cameron, Gavin Megaw, Raymond Bond, Maurice Mulvenna, Siobhan O’Neill, Cherie Armour, and Michael McTear. Assessing the usability of a chatbot for mental health care. In Internet Science: INSCI 2018 International Workshops, St. Petersburg, Russia, Octobe...

2018
[20]

Cdc wonder - multiple cause of death, 1999-2022, 2022

Centers for Disease Control and Prevention. Cdc wonder - multiple cause of death, 1999-2022, 2022

1999
[21]

Can language models be instructed to protect personal information? arXiv preprint arXiv:2310.02224, 2023

Yang Chen, Ethan Mendes, Sauvik Das, Wei Xu, and Alan Ritter. Can language models be instructed to protect personal information? arXiv preprint arXiv:2310.02224, 2023

2023 arXiv
[22]

What is the im- pact of mental health-related stigma on help-seeking? a systematic review of quantitative and qualitative studies.Psychological medicine, 45(1):11–27, 2015

Sarah Clement, Oliver Schauman, Tanya Graham, Francesca Mag- gioni, Sara Evans-Lacko, Nikita Bezborodovs, Craig Morgan, Nicolas Rüsch, June SL Brown, and Graham Thornicroft. What is the im- pact of mental health-related stigma on help-seeking? a systematic review of quantitati...

2015
[23]

Be- lieving anthropomorphism: Examining the role of anthropomorphic cues on trust in large language models

Michelle Cohn, Mahima Pushkarna, Gbolahan O Olanubi, Joseph M Moran, Daniel Padgett, Zion Mengesha, and Courtney Heldreth. Be- lieving anthropomorphism: Examining the role of anthropomorphic cues on trust in large language models. In Extended Abstracts of the CHI Conference on...

2024
[24]

Basics of qualitative research , volume 14

Juliet Corbin and Anselm Strauss. Basics of qualitative research , volume 14. sage, 2015

2015
[25]

How stigma interferes with mental health care

Patrick Corrigan. How stigma interferes with mental health care. American psychologist, 59(7):614, 2004

2004
[26]

Understanding the impact of stigma on people with mental illness

Patrick W Corrigan and Amy C Watson. Understanding the impact of stigma on people with mental illness. World psychiatry, 1(1):16, 2002

2002
[27]

Benefits and harms of large language models in digital mental health

Munmun De Choudhury, Sachin R Pendse, and Neha Kumar. Benefits and harms of large language models in digital mental health. arXiv preprint arXiv:2311.14693, 2023

2023 arXiv
[28]

The health risks of generative ai-based wellness apps

Julian De Freitas and I Glenn Cohen. The health risks of generative ai-based wellness apps. Nature Medicine, pages 1–7, 2024

2024
[29]

Assess- ing prognosis in depression: comparing perspectives of ai models, mental health professionals and the general public

Zohar Elyoseph, Inbar Levkovich, and Shiri Shinan-Altman. Assess- ing prognosis in depression: comparing perspectives of ai models, mental health professionals and the general public. Family Medicine and Community Health, 12, 2024

2024
[30]

The empatica health monitoring platform receives fda clearance, 2022

Empatica. The empatica health monitoring platform receives fda clearance, 2022

2022
[31]

De- livering cognitive behavior therapy to young adults with symptoms of depression and anxiety using a fully automated conversational agent (woebot): a randomized controlled trial

Kathleen Kara Fitzpatrick, Alison Darcy, and Molly Vierhile. De- livering cognitive behavior therapy to young adults with symptoms of depression and anxiety using a fully automated conversational agent (woebot): a randomized controlled trial. JMIR mental health, 4(2):e7785, 2017

2017
[32]

General wellness: Policy for low risk devices

Center for Devices, US Food Radiological health, and Drug Admin. General wellness: Policy for low risk devices
[33]

Regulating mass surveillance as privacy pollu- tion: Learning from environemntal impact statements

A Michael Froomkin. Regulating mass surveillance as privacy pollu- tion: Learning from environemntal impact statements. U. Ill. L. Rev., page 1713, 2015

2015
[34]

Using psychological artificial intelligence (tess) to relieve symptoms of depression and anxiety: randomized controlled trial

Russell Fulmer, Angela Joerin, Breanna Gentile, Lysanne Lakerink, Michiel Rauws, et al. Using psychological artificial intelligence (tess) to relieve symptoms of depression and anxiety: randomized controlled trial. JMIR mental health, 5(4):e9782, 2018

2018
[35]

Early experiences with e-health services (1999–2002): promise, reality, and implications

Vicki Fung, Eduardo Ortiz, Jie Huang, Bruce Fireman, Robert Miller, Joseph V Selby, and John Hsu. Early experiences with e-health services (1999–2002): promise, reality, and implications. Medical care, 44(5):491–496, 2006

1999
[36]

Basing cybersecurity training on user perceptions

Susanne Furman, Mary Frances Theofanos, Yee-Yin Choong, and Brian Stanton. Basing cybersecurity training on user perceptions. IEEE Security & Privacy, 10(2):40–49, 2011

2011
[37]

Retrieval-augmented generation for large language models: A survey

Yunfan Gao, Yun Xiong, Xinyu Gao, Kangxiang Jia, Jinliu Pan, Yuxi Bi, Yi Dai, Jiawei Sun, and Haofen Wang. Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997, 2023

2023 arXiv
[38]

Looking for trouble: under- standing end-user security management

Joshua B Gross and Mary Beth Rosson. Looking for trouble: under- standing end-user security management. In Proceedings of the 2007 Symposium on Computer Human interaction For the Management of information Technology, pages 10–es, 2007

2007
[39]

it’s the company, the government, you and i

Julie Haney, Yasemin Acar, and Susanne Furman. " it’s the company, the government, you and i": User perceptions of responsibility for smart home privacy and security. In30th USENIX Security Symposium (USENIX Security 21), pages 411–428, 2021

2021
[40]

what can i really do?

Eszter Hargittai and Alice Marwick. “what can i really do?” explaining the privacy paradox with online apathy. International journal of communication, 10:21, 2016

2016
[41]

Attention is not all you need: the complicated case of ethically using large language models in healthcare and medicine

Stefan Harrer. Attention is not all you need: the complicated case of ethically using large language models in healthcare and medicine. EBioMedicine, 90, 2023

2023
[42]

Physician versus large language model chatbot re- sponses to web-based questions from autistic patients in chinese: Cross-sectional comparative analysis

Wenjie He, Wenyan Zhang, Ya Jin, Qiang Zhou, Huadan Zhang, and Qing Xia. Physician versus large language model chatbot re- sponses to web-based questions from autistic patients in chinese: Cross-sectional comparative analysis. Journal of Medical Internet Research, 26:e54706, 2024

2024
[43]

Conversational agent interventions for mental health problems: systematic review and meta-analysis of ran- domized controlled trials

Yuhao He, Li Yang, Chunlian Qian, Tong Li, Zhengyuan Su, Qiang Zhang, and Xiangqing Hou. Conversational agent interventions for mental health problems: systematic review and meta-analysis of ran- domized controlled trials. Journal of medical Internet research , 25:e43862, 2023

2023
[44]

A paradigm shift: Consumer attitudes toward mental health technology in 2021, 2021

Woebot Health. A paradigm shift: Consumer attitudes toward mental health technology in 2021, 2021

2021
[45]

Designated health professional shortage areas statistics

Health Resources and Services Administration. Designated health professional shortage areas statistics. https://data.hrsa.gov/ topics/health-workforce/shortage-areas, 2025

2025
[46]

Depres- sion and decision-making capacity for treatment or research: a sys- tematic review

Thomas Hindmarch, Matthew Hotopf, and Gareth S Owen. Depres- sion and decision-making capacity for treatment or research: a sys- tematic review. BMC medical ethics, 14:1–10, 2013

2013
[47]

Psychological, re- lational, and emotional effects of self-disclosure after conversations with a chatbot

Annabell Ho, Jeff Hancock, and Adam S Miner. Psychological, re- lational, and emotional effects of self-disclosure after conversations with a chatbot. Journal of Communication, 68(4):712–733, 2018

2018
[48]

Pri- vacy cynicism: A new approach to the privacy paradox

Christian Pieter Hoffmann, Christoph Lutz, and Giulia Ranzini. Pri- vacy cynicism: A new approach to the privacy paradox. Cyberpsy- chology: Journal of Psychosocial Research on Cyberspace , 10(4), 2016

2016
[49]

Challenges in build- ing intelligent open-domain dialog systems

Minlie Huang, Xiaoyan Zhu, and Jianfeng Gao. Challenges in build- ing intelligent open-domain dialog systems. ACM Transactions on Information Systems (TOIS), 38(3):1–32, 2020

2020
[50]

Do you trust chatgpt?–perceived credibility of human and ai- generated content

Martin Huschens, Martin Briesch, Dominik Sobania, and Franz Roth- lauf. Do you trust chatgpt?–perceived credibility of human and ai- generated content. arXiv preprint arXiv:2309.02524, 2023

2023 arXiv
[51]

Internet services for communicating with the general practice: barely noticed and used by patients

Martine WJ Huygens, Joan Vermeulen, Roland D Friele, Onno CP van Schayck, Judith D de Jong, and Luc P de Witte. Internet services for communicating with the general practice: barely noticed and used by patients. Interactive journal of medical research, 4(4):e4245, 2015

2015
[52]

An empathy-driven, conversational artificial intelligence agent (wysa) for digital mental well-being: real-world data evaluation mixed-methods study

Becky Inkster, Shubhankar Sarda, Vinod Subramanian, et al. An empathy-driven, conversational artificial intelligence agent (wysa) for digital mental well-being: real-world data evaluation mixed-methods study. JMIR mHealth and uHealth, 6(11):e12106, 2018

2018
[53]

Privacy concerns in chatbot interactions

Carolin Ischen, Theo Araujo, Hilde V oorveld, Guda van Noort, and Edith Smit. Privacy concerns in chatbot interactions. In Chatbot Research and Design: Third International Workshop, CONVERSA- TIONS 2019, Amsterdam, The Netherlands, November 19–20, 2019, Revised Selected Papers...

2019
[54]

The gramm-leach-bliley act, information privacy, and the limits of default rules

Edward J Janger and Paul M Schwartz. The gramm-leach-bliley act, information privacy, and the limits of default rules. Minn. L. Rev., 86:1219, 2001

2001
[55]

A chatbot was designed to help prevent eating disorders

Julie Jargon. A chatbot was designed to help prevent eating disorders. then it gave dieting tips. The Wall Street Journal, 2023

2023
[56]

Health care privacy risks of ai chatbots

Genevieve P Kanter and Eric A Packel. Health care privacy risks of ai chatbots. JAMA, 2023

2023
[57]

Chatgpt for good? on opportunities and challenges of large language models for education

Enkelejda Kasneci, Kathrin Seßler, Stefan Küchemann, Maria Ban- nert, Daryna Dementieva, Frank Fischer, Urs Gasser, Georg Groh, Stephan Günnemann, Eyke Hüllermeier, et al. Chatgpt for good? on opportunities and challenges of large language models for education. Learning and in...

2023
[58]

Mindfuldiary: Harnessing large language model to support psychiatric patients’ journaling

Taewan Kim, Seolyeong Bae, Hyun Ah Kim, Su-Woo Lee, Hwajung Hong, Chanmo Yang, and Young-Ho Kim. Mindfuldiary: Harnessing large language model to support psychiatric patients’ journaling. In Proceedings of the 2024 CHI Conference on Human Factors in Com- puting Systems, CHI ’2...

2024
[59]

Anthropomorphism of com- puters: Is it mindful or mindless? Computers in Human Behavior , 28(1):241–250, 2012

Youjeong Kim and S Shyam Sundar. Anthropomorphism of com- puters: Is it mindful or mindless? Computers in Human Behavior , 28(1):241–250, 2012

2012
[60]

From promise to practice: towards the realisation of ai-informed mental health care

Nikolaos Koutsouleris, Tobias U Hauser, Vasilisa Skvortsova, and Munmun De Choudhury. From promise to practice: towards the realisation of ai-informed mental health care. The Lancet Digital Health, 4(11):e829–e840, 2022

2022
[61]

mental health crisis

Bridget M Kuehn. Clinician shortage exacerbates pandemic-fueled “mental health crisis”. JAMA, 327(22):2179–2181, 2022

2022
[62]

The opportunities and risks of large language models in mental health

Hannah R Lawrence, Renee A Schneider, Susan B Rubin, Maja J Matari´c, Daniel J McDuff, and Megan Jones Bell. The opportunities and risks of large language models in mental health. JMIR Mental Health, 11(1):e59479, 2024

2024
[63]

Designing a chatbot as a mediator for promoting deep self-disclosure to a real mental health professional

Yi-Chieh Lee, Naomi Yamashita, and Yun Huang. Designing a chatbot as a mediator for promoting deep self-disclosure to a real mental health professional. Proceedings of the ACM on Human-Computer Interaction, 4(CSCW1):1–27, 2020

2020
[64]

i hear you, i feel you

Yi-Chieh Lee, Naomi Yamashita, Yun Huang, and Wai Fu. " i hear you, i feel you": encouraging deep self-disclosure through a chatbot. In Proceedings of the 2020 CHI conference on human factors in computing systems, pages 1–12, 2020

2020
[65]

User perception of wysa as a mental well-being support tool during the covid-19 pandemic

Carlos Miguel Legaspi Jr, Tristan Raphael Pacana, Kyle Loja, Christina Sing, and Ethel Ong. User perception of wysa as a mental well-being support tool during the covid-19 pandemic. InProceedings of the Asian HCI Symposium 2022, pages 52–57, 2022

2022
[66]

Systematic review and meta-analysis of ai-based conversational agents for promoting mental health and well-being

Han Li, Renwen Zhang, Yi-Chieh Lee, Robert E Kraut, and David C Mohr. Systematic review and meta-analysis of ai-based conversational agents for promoting mental health and well-being. NPJ Digital Medicine, 6(1):236, 2023

2023
[67]

Harness- ing large language models’ empathetic response generation capabil- ities for online mental health counselling support

Siyuan Brandon Loh and Aravind Sesagiri Raamkumar. Harness- ing large language models’ empathetic response generation capabil- ities for online mental health counselling support. arXiv preprint arXiv:2310.08017, 2023

2023 arXiv
[68]

It’s only a computer: Virtual humans increase willingness to disclose

Gale M Lucas, Jonathan Gratch, Aisha King, and Louis-Philippe Morency. It’s only a computer: Virtual humans increase willingness to disclose. Computers in Human Behavior, 37:94–100, 2014

2014
[69]

A fully auto- mated conversational agent for promoting mental well-being: a pilot rct using mixed methods

Kien Hoa Ly, Ann-Marie Ly, and Gerhard Andersson. A fully auto- mated conversational agent for promoting mental well-being: a pilot rct using mixed methods. Internet interventions, 10:39–46, 2017

2017
[70]

Understanding the benefits and challenges of using large language model-based conversational agents for mental well-being support

Zilin Ma, Yiyang Mei, and Zhaoyuan Su. Understanding the benefits and challenges of using large language model-based conversational agents for mental well-being support. In AMIA Annual Symposium Proceedings, volume 2023, page 1105. American Medical Informatics Association, 2023

2023
[71]

Evalu- ating user feedback for an artificial intelligence–enabled, cognitive behavioral therapy–based mental health app (wysa): qualitative the- matic analysis

Tanya Malik, Adrian Jacques Ambrose, Chaitali Sinha, et al. Evalu- ating user feedback for an artificial intelligence–enabled, cognitive behavioral therapy–based mental health app (wysa): qualitative the- matic analysis. JMIR Human Factors, 9(2):e35668, 2022

2022
[72]

Ai chatbots, health privacy, and challenges to hipaa compliance

Mason Marks and Claudia E Haupt. Ai chatbots, health privacy, and challenges to hipaa compliance. Jama, 2023

2023
[73]

Will increased disclosure help-evaluating the recommendations of the ali’s principles of the law of software contracts

Florencia Marotta-Wurgler. Will increased disclosure help-evaluating the recommendations of the ali’s principles of the law of software contracts. U. Chi. L. Rev., 78:165, 2011

2011
[74]

Privacy notices as tabula rasa: An empirical investi- gation into how complying with a privacy notice is related to meeting privacy expectations online

Kirsten Martin. Privacy notices as tabula rasa: An empirical investi- gation into how complying with a privacy notice is related to meeting privacy expectations online. Journal of Public Policy & Marketing, 34(2):210–227, 2015

2015
[75]

Measuring privacy: An empir- ical test using context to expose confounding variables

Kirsten Martin and Helen Nissenbaum. Measuring privacy: An empir- ical test using context to expose confounding variables. Colum. Sci. & Tech. L. Rev., 18:176, 2016

2016
[76]

The cost of reading privacy policies

Aleecia M McDonald and Lorrie Faith Cranor. The cost of reading privacy policies. Isjlp, 4:543, 2008

2008
[77]

The law of friction

William McGeveran. The law of friction. U. Chi. Legal F ., page 15, 2013

2013
[78]

Mental health america: Adult data 2024, 2024

Mental Health America. Mental health america: Adult data 2024, 2024

2024
[79]

Strategies for reducing online privacy risks: Why consumers read (or don’t read) online privacy notices

George R Milne and Mary J Culnan. Strategies for reducing online privacy risks: Why consumers read (or don’t read) online privacy notices. Journal of interactive marketing, 18(3):15–29, 2004

2004
[80]

Challenges and opportunities to meet the mental health needs of un- derserved and disenfranchised populations in the united states

Francesca Mongelli, Penelope Georgakopoulos, and Michele T Pato. Challenges and opportunities to meet the mental health needs of un- derserved and disenfranchised populations in the united states. Focus, 18(1):16–24, 2020

2020
[81]

Institute of museum and library services funding cuts under trump order, 2025

Nadia Lathan. Institute of museum and library services funding cuts under trump order, 2025

2025
[82]

Privacy as contextual integrity

Helen Nissenbaum. Privacy as contextual integrity. Wash. L. Rev., 79:119, 2004

2004
[83]

Effectiveness of a conversational chatbot (dejal@ bot) for the adult population to quit smoking: pragmatic, multicenter, controlled, randomized clinical trial in primary care

Eduardo Olano-Espinosa, Jose Francisco Avila-Tomas, Cesar Minue- Lorenzo, Blanca Matilla-Pardo, María Encarnación Serrano Serrano, F Javier Martinez-Suberviola, Mario Gil-Conesa, Isabel Del Cura- González, et al. Effectiveness of a conversational chatbot (dejal@ bot) for the a...

2022
[84]

Omiye, Jenna Lester, Simon Spichak, Veronica Rotem- berg, and Roxana Daneshjou

Jesutofunmi A. Omiye, Jenna Lester, Simon Spichak, Veronica Rotem- berg, and Roxana Daneshjou. Beyond the hype: large language models propagate race-based medicine. medRxiv, 2023

2023
[85]

Privacy policy, 2024

OpenAI. Privacy policy, 2024

2024
[86]

Demand characteristics and the concept of quasi- controls

Martin T Orne. Demand characteristics and the concept of quasi- controls. Artifacts in behavioral research: Robert Rosenthal and Ralph L. Rosnow’s classic books, 110:110–137, 2009

2009
[87]

Massive data language models and conversa- tional artificial intelligence: Emerging issues

Daniel E O’Leary. Massive data language models and conversa- tional artificial intelligence: Emerging issues. Intelligent Systems in Accounting, Finance and Management, 29(3):182–198, 2022

2022
[88]

Beneficent dehumanization: Employing artificial intelligence and carebots to mitigate shame- induced barriers to medical care

Amitabha Palmer and David Schwan. Beneficent dehumanization: Employing artificial intelligence and carebots to mitigate shame- induced barriers to medical care. Bioethics, 36(2):187–193, 2022

2022
[89]

Artificial intelligence and algorithmic bias: implications for health systems

Trishan Panch, Heather Mattie, and Rifat Atun. Artificial intelligence and algorithmic bias: implications for health systems. Journal of global health, 9(2), 2019

2019
[90]

i wrote as if i were telling a story to some- one i knew

SoHyun Park, Anja Thieme, Jeongyun Han, Sungwoo Lee, Wonjong Rhee, and Bongwon Suh. “i wrote as if i were telling a story to some- one i knew.”: Designing chatbot interactions for expressive writing in mental health. In Proceedings of the 2021 ACM Designing Interactive Systems...

2021
[91]

Perlis, Joseph F Goldberg, Michael Joshua Ostacher, and Christopher D Schneck

Roy H. Perlis, Joseph F Goldberg, Michael Joshua Ostacher, and Christopher D Schneck. Clinical decision support for bipolar de- pression using large language models. Neuropsychopharmacology, 49:1412 – 1416, 2024

2024
[92]

A therapeutic relational agent for reducing problematic substance use (woebot): development and usability study.Journal of medical Internet research, 23(3):e24850, 2021

Judith J Prochaska, Erin A V ogel, Amy Chieng, Matthew Kendra, Michael Baiocchi, Sarah Pajarito, and Athena Robinson. A therapeutic relational agent for reducing problematic substance use (woebot): development and usability study.Journal of medical Internet research, 23(3):e24...

2021
[93]

Privacy policy, 2024

Replika. Privacy policy, 2024

2024
[94]

A protection motivation theory of fear appeals and attitude change1

Ronald W Rogers. A protection motivation theory of fear appeals and attitude change1. The journal of psychology, 91(1):93–114, 1975

1975
[95]

The coding manual for qualitative researchers

Johnny Saldaña. The coding manual for qualitative researchers. sage, 2021

2021
[96]

Pocket skills: A conversational mobile web app to support dialectical behavioral therapy

Jessica Schroeder, Chelsey Wilkes, Kael Rowan, Arturo Toledo, Ann Paradiso, Mary Czerwinski, Gloria Mark, and Marsha M Linehan. Pocket skills: A conversational mobile web app to support dialectical behavioral therapy. In Proceedings of the 2018 CHI Conference on Human Factors ...

2018
[97]

Epistemic injustice and mental illness

Anastasia Philippa Scrutton. Epistemic injustice and mental illness. In The Routledge handbook of epistemic injustice , pages 347–355. Routledge, 2017

2017
[98]

Clinical accuracy of large language models and google search responses to postpartum depression questions: cross-sectional study

Emre Sezgin, Faraaz Chekeni, Jennifer Lee, and Sarah Keim. Clinical accuracy of large language models and google search responses to postpartum depression questions: cross-sectional study. Journal of Medical Internet Research, 25:e49240, 2023

2023
[99]

Cognitive reframing of negative thoughts through human-language model interaction

Ashish Sharma, Kevin Rushton, Inna Wanyin Lin, David Wadden, Khendra G Lucas, Adam S Miner, Theresa Nguyen, and Tim Althoff. Cognitive reframing of negative thoughts through human-language model interaction. arXiv preprint arXiv:2305.02466, 2023

2023 arXiv
[100]

Mental health stigma update: A review of consequences.Advances in Mental Health, 12(3):202–215, 2014

Amy E Sickel, Jason D Seacat, and Nina A Nabors. Mental health stigma update: A review of consequences.Advances in Mental Health, 12(3):202–215, 2014

2014
[101]

it just happened to be the perfect thing

Steven Siddals, Astrid Coxon, and John Torous. " it just happened to be the perfect thing": Real-life experiences of generative ai chatbots for mental health. 2024

2024
[102]

Skating the line between general wellness products and regulated devices: strategies and implications

David A Simon, Carmel Shachar, and I Glenn Cohen. Skating the line between general wellness products and regulated devices: strategies and implications. Journal of Law and the Biosciences, 9(2):lsac015, 2022

2022
[103]

Virtual reality job interview training in adults with autism spectrum disorder

Matthew J Smith, Emily J Ginger, Katherine Wright, Michael A Wright, Julie Lounds Taylor, Laura Boteler Humm, Dale E Olsen, Morris D Bell, and Michael F Fleming. Virtual reality job interview training in adults with autism spectrum disorder. Journal of autism and developmental...

2014
[104]

Lower lev- els of directed exploration and reflective thinking are associated with greater anxiety and depression

Ryan Smith, Samuel Taylor, Robert C Wilson, Anne E Chuning, Michelle R Persich, Siyu Wang, and William DS Killgore. Lower lev- els of directed exploration and reflective thinking are associated with greater anxiety and depression. Frontiers in Psychiatry, 12:782136, 2022

2022
[105]

Guided versus unguided chatbot-delivered cognitive behav- ioral intervention for individuals with moderate-risk and problem gambling: A randomized controlled trial (gambot2 study)

Ryuhei So, Naoki Emura, Kozue Okazaki, Sakiko Takeda, Takashi Sunami, Kohei Kitagawa, Yoshitake Takebayashi, and Toshi A Fu- rukawa. Guided versus unguided chatbot-delivered cognitive behav- ioral intervention for individuals with moderate-risk and problem gambling: A randomiz...

2024
[106]

The myth of the privacy paradox

Daniel J Solove. The myth of the privacy paradox. Geo. Wash. L. Rev., 89:1, 2021

2021
[107]

Pendse, Neha Kumar, and Munmun De Choud- hury

Inhwa Song, Sachin R. Pendse, Neha Kumar, and Munmun De Choud- hury. The Typing Cure: Experiences with Large Language Model Chatbots for Mental Health Support, March 2024

2024
[108]

E-privacy in 2nd generation e-commerce: privacy preferences versus actual behavior

Sarah Spiekermann, Jens Grossklags, and Bettina Berendt. E-privacy in 2nd generation e-commerce: privacy preferences versus actual behavior. In Proceedings of the 3rd ACM conference on Electronic Commerce, pages 38–47, 2001

2001
[109]

User experiences of social support from companion chatbots in everyday contexts: thematic analysis

Vivian Ta, Caroline Griffith, Carolynn Boatfield, Xinyu Wang, Maria Civitello, Haley Bader, Esther DeCero, Alexia Loggarakis, et al. User experiences of social support from companion chatbots in everyday contexts: thematic analysis. Journal of medical Internet research , 22(3)...

2020
[110]

A general inductive approach for analyzing qualita- tive evaluation data

David R Thomas. A general inductive approach for analyzing qualita- tive evaluation data. American journal of evaluation, 27(2):237–246, 2006

2006
[111]

Belgian man commits suicide following ex- changes with chatbot, 2023

The Brussels Times. Belgian man commits suicide following ex- changes with chatbot, 2023

2023
[112]

Can you see me now? audience and disclosure regu- lation in online social network sites

Zeynep Tufekci. Can you see me now? audience and disclosure regu- lation in online social network sites. Bulletin of Science, Technology & Society, 28(1):20–36, 2008

2008
[113]

Open to exploitation: American shoppers online and offline, 2005

Joseph Turow, Lauren Feldman, and Kimberly Meltzer. Open to exploitation: American shoppers online and offline, 2005

2005
[114]

Americans reject tailored advertising and three activities that enable it

Joseph Turow, Jennifer King, Chris Jay Hoofnagle, Amy Bleakley, and Michael Hennessy. Americans reject tailored advertising and three activities that enable it. Available at SSRN 1478214, 2009

2009
[115]

Chatbots and conversa- tional agents in mental health: a review of the psychiatric landscape

Aditya Nrusimha Vaidyam, Hannah Wisniewski, John David Halamka, Matcheri S Kashavan, and John Blake Torous. Chatbots and conversa- tional agents in mental health: a review of the psychiatric landscape. The Canadian Journal of Psychiatry, 64(7):456–464, 2019

2019
[116]

Perceived public stigma and the willingness to seek counseling: The mediat- ing roles of self-stigma and attitudes toward counseling

David L V ogel, Nathaniel G Wade, and Ashley H Hackler. Perceived public stigma and the willingness to seek counseling: The mediat- ing roles of self-stigma and attitudes toward counseling. Journal of counseling psychology, 54(1):40, 2007

2007
[117]

The mind in the ma- chine: Anthropomorphism increases trust in an autonomous vehicle

Adam Waytz, Joy Heafner, and Nicholas Epley. The mind in the ma- chine: Anthropomorphism increases trust in an autonomous vehicle. Journal of experimental social psychology, 52:113–117, 2014

2014
[118]

Ethical and social risks of harm from language models

Laura Weidinger, John Mellor, Maribeth Rauh, Conor Griffin, Jonathan Uesato, Po-Sen Huang, Myra Cheng, Mia Glaese, Borja Balle, Atoosa Kasirzadeh, et al. Ethical and social risks of harm from language models. arXiv preprint arXiv:2112.04359, 2021

2021 arXiv
[119]

Are racial and ethnic minorities less willing to participate in health research? PLoS medicine, 3(2):e19, 2006

David Wendler, Raynard Kington, Jennifer Madans, Gretchen Van Wye, Heidi Christ-Schmidt, Laura A Pratt, Otis W Brawley, Cary P Gross, and Ezekiel Emanuel. Are racial and ethnic minorities less willing to participate in health research? PLoS medicine, 3(2):e19, 2006

2006
[120]

The shaky foundations of large language models and foundation models for electronic health records

Michael Wornow, Yizhe Xu, Rahul Thapa, Birju Patel, Ethan Stein- berg, Scott Fleming, Michael A Pfeffer, Jason Fries, and Nigam H Shah. The shaky foundations of large language models and foundation models for electronic health records. npj Digital Medicine, 6(1):135, 2023

2023
[121]

‘he would still be here’: Man dies by suicide after talking with ai chatbot, widow says

Chloe Xiang. ‘he would still be here’: Man dies by suicide after talking with ai chatbot, widow says. Vice, 2023

2023
[122]

Rodriguez, Leo An- thony Celi, Judy Gichoya, Dan Jurafsky, Peter Szolovits, David W

Travis Zack, Eric Lehman, Mirac Suzgun, Jorge A. Rodriguez, Leo An- thony Celi, Judy Gichoya, Dan Jurafsky, Peter Szolovits, David W. Bates, Raja-Elie E. Abdulnour, Atul J. Butte, and Emily Alsentzer. Coding inequity: Assessing gpt-4’s potential for perpetuating racial and gen...

2023
[123]

Popularity of Mental Health Chatbots Grows

Nick Zagorski. Popularity of Mental Health Chatbots Grows. Psychi- atric News, 57(5), May 2022

2022
[124]

The artificial intelligence large language models and neuropsychiatry practice and research ethic

Yi Zhong, Yu-jun Chen, Yang Zhou, Jia-Jun Yin, Yu-jun Gao, et al. The artificial intelligence large language models and neuropsychiatry practice and research ethic. Asian journal of psychiatry, 84:103577, 2023

2023
[125]

In- ducing positive perspectives with text reframing

Caleb Ziems, Minzhi Li, Anthony Zhang, and Diyi Yang. In- ducing positive perspectives with text reframing. arXiv preprint arXiv:2204.02952, 2022

2022 arXiv
[126]

Anthropomorphism: opportunities and chal- lenges in human–robot interaction

Jakub Złotowski, Diane Proudfoot, Kumar Yogeeswaran, and Christoph Bartneck. Anthropomorphism: opportunities and chal- lenges in human–robot interaction. International journal of social robotics, 7:347–360, 2015. Available Artifacts Attached to the following link we make avail...

2015 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

An overview of the features of chatbots in mental health: A scoping review

Alaa A Abd-Alrazaq, Mohannad Alajlani, Ali Abdallah Alalwan, Brid- gette M Bewick, Peter Gardner, and Mowafa Househ. An overview of the features of chatbots in mental health: A scoping review. Inter- national journal of medical informatics, 132:103978, 2019

2019

[2] [2]

Perceptions and opinions of patients about mental health chatbots: scoping review

Alaa A Abd-Alrazaq, Mohannad Alajlani, Nashva Ali, Kerstin De- necke, Bridgette M Bewick, and Mowafa Househ. Perceptions and opinions of patients about mental health chatbots: scoping review. Journal of medical Internet research, 23(1):e17828, 2021

2021

[3] [3]

Large language models associate muslims with violence.Nature Machine Intelligence, 3:461–463, 06 2021

Abubakar Abid, Maheen Farooqi, and James Zou. Large language models associate muslims with violence.Nature Machine Intelligence, 3:461–463, 06 2021

2021

[4] [4]

Imagined communities: Aware- ness, information sharing, and privacy on the facebook

Alessandro Acquisti and Ralph Gross. Imagined communities: Aware- ness, information sharing, and privacy on the facebook. In Interna- tional workshop on privacy enhancing technologies , pages 36–58. Springer, 2006

2006

[5] [5]

Privacy attitudes and pri- vacy behavior: Losses, gains, and hyperbolic discounting

Alessandro Acquisti and Jens Grossklags. Privacy attitudes and pri- vacy behavior: Losses, gains, and hyperbolic discounting. In Eco- nomics of information security, pages 165–178. Springer, 2004

2004

[6] [6]

Privacy and rationality in individual decision making

Alessandro Acquisti and Jens Grossklags. Privacy and rationality in individual decision making. IEEE security & privacy , 3(1):26–33, 2005

2005

[7] [7]

What can behavioral eco- nomics teach us about privacy? In Digital privacy, pages 363–378

Alessandro Acquisti and Jens Grossklags. What can behavioral eco- nomics teach us about privacy? In Digital privacy, pages 363–378. Auerbach Publications, 2007

2007

[8] [8]

Assessing the effectiveness of chatgpt in delivering mental health support: a qualitative study.Journal of Multidisciplinary Healthcare, pages 461–471, 2024

Fahad Alanezi. Assessing the effectiveness of chatgpt in delivering mental health support: a qualitative study.Journal of Multidisciplinary Healthcare, pages 461–471, 2024

2024