Evaluating the Clinical Safety of LLMs in Response to High-Risk Mental Health Disclosures

Siddharth Shah; Amit Gupta; Aarav Mann; Alexandre Vaz; Benjamin E. Caldwell; Robert Scholz; Peter Awad; Rocky Allemandi; Doug Faust; Harshita Banka; Tony Rousmaniere

arXiv · 2509.08839

Evaluating the Clinical Safety of LLMs in Response to High-Risk Mental Health Disclosures

Siddharth Shah·Amit Gupta·Aarav Mann·Alexandre Vaz·Benjamin E. Caldwell·Robert Scholz·Peter Awad·Rocky Allemandi·Doug Faust·Harshita Banka·Tony Rousmaniere

View original source ↗Read source document ↗Export metadata ↓

Abstract

As large language models (LLMs) increasingly mediate emotionally sensitive conversations, especially in mental health contexts, their ability to recognize and respond to high-risk situations becomes a matter of public safety. This study evaluates the responses of six popular LLMs (Claude, Gemini, Deepseek, ChatGPT, Grok 3, and LLAMA) to user prompts simulating crisis-level mental health disclosures. Drawing on a coding framework developed by licensed clinicians, five safety-oriented behaviors were assessed: explicit risk acknowledgment, empathy, encouragement to seek help, provision of specific resources, and invitation to continue the conversation. Claude outperformed all others in global assessment, while Grok 3, ChatGPT, and LLAMA underperformed across multiple domains. Notably, most models exhibited empathy, but few consistently provided practical support or sustained engagement. These findings suggest that while LLMs show potential for emotionally attuned communication, none currently meet satisfactory clinical standards for crisis response. Ongoing development and targeted fine-tuning are essential to ensure ethical deployment of AI in mental health settings.

Explore related subjects

cs.CY

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Siddharth Shah, Amit Gupta, Aarav Mann, Alexandre Vaz, Benjamin E. Caldwell, Robert Scholz, Peter Awad, Rocky Allemandi, Doug Faust, Harshita Banka, Tony Rousmaniere. 2025-09-01. Evaluating the Clinical Safety of LLMs in Response to High-Risk Mental Health Disclosures. https://arxiv.org/abs/2509.08839

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

An Investigation Into Secondary School Students' Debugging Behaviour in Python

Background and context: Debugging is a significant and often frustrating challenge for beginner programmers. Understanding students' debugging behaviours and strategies can help to identify common difficulties and inform approaches for alleviating these. Currently, there are limited studies of school students' debugging behaviour in a text-based programming language, a medium through which millions are learning to program. Objectives: In this paper, we investigate the debugging behaviour of 12-14-year-old students learning Python through a lesson-long classroom study. Method: We collected program snapshots from 73 students' attempts at a set of Python debugging exercises in an online code editor. Through qualitative content analysis of these snapshots, we developed a granular categorisation of the changes students made when debugging. Findings: A range of debugging behaviours were exhibited by students, many of which were ineffective. Students added errors through small-scale changes, reverted corrective changes, and repeatedly ran identical programs in quick succession. From the results, we identify four barriers to successful and reliable debugging for students learning a text-based programming language: fragile knowledge, a lack of systematicity and reflection, the syntax barrier, and dynamics of emotions and attitudes. Implications: This paper highlights some of the difficulties that secondary school students have when debugging in Python and the challenges of analysing program snapshots through manual inspection. We recommend that school teachers explicitly teach a systematic approach to debugging and discourage the use of ineffective debugging behaviours, and that programming environments should contain features that facilitate successful debugging.

cs.CY↗

arXiv2026

Printed but not benchmarkable: most building-decarbonisation disclosure cannot be matched to the pathways that stranding regulation assumes

Cities are beginning to enforce carbon limits on existing buildings. Science-based decarbonisation pathways set those limits one asset type and one jurisdiction at a time. Owners, however, report for the whole firm. We measure what that mismatch costs on two sets of public corporate reports: a census of 502 reports from the 119 listed built-environment firms with a collected report inside a 2,246-firm panel (2003-2023), and 519 real-estate reports from 101 firms (2007-2024). BeDA, a multimodal language-model tool whose reliability we test first, read them. Running the pathway frameworks' own entry tests over published disclosure: 16.5% of census reports (43.7% of real-estate reports) print an operational carbon intensity per square metre; 6.6% (25.0%) can be matched to a pathway for their property type in a covered jurisdiction; and only 5.0% (16.4%) disclose the floor area they divided by. Of the failures at the pathway test, 82-84% follow from reports lumping the portfolio together and 16-18% from a missing curve in the pathway library. The obstacle is the reporting unit, not missing data. The rate is roughly twice as high for European as for US listings (65-71% versus 35% in listed real estate). We also show that a US portfolio's carbon verdict cannot be worked out from disclosure at all. Within one climate zone, the pathway's carbon limit varies by up to 2.79-fold with the electricity subregion, which no report names; its energy limit does not move. Extraction is checked against the source PDFs (97.3% of extracted intensities appear verbatim) and repeats on a second extractor (kappa = 0.97). Recall of the non-disclosing class was 95.1% in a blinded hand audit of 122 reports. The fix follows from the measurement: split intensity by asset type and jurisdiction, and report floor area.

cs.CY↗

arXiv2026

Incipit: Axiom-Grounded Scaffolding for Human-AI Literary Creation

Large language models can produce fluent prose from short prompts, but a direct interaction gives writers little access to the assumptions that shape a long narrative. We present Incipit, an implemented research prototype that inserts an explicit planning layer between writer intent and generated prose. The layer is grounded in literary axioms, which are curated and reusable propositions about human experience and narrative craft. The prototype connects a knowledge base of 1455 axioms and 472 typed relationships to a five round direction dialogue, a retrieval and selection pipeline, and a three level blueprint covering creative premises, story beats, character arcs, and chapter outlines. Writers can inspect and edit the resulting structures before using them as context for scene generation. Additional modules support real event abstraction and five dimensional diagnostic feedback. We describe the design rationale, data flow, implementation boundaries, and a worked design example. Because no controlled user study or independently rated output study has yet been completed, we do not claim that the system improves literary quality. Instead, we outline a future preregistered comparison designed to distinguish the contribution of axiom grounding from that of hierarchical planning. The paper contributes a concrete architecture for making literary knowledge an inspectable coordination object in human AI writing.

cs.CY↗