About Metonym Clinical AI Intelligence
Metonym Clinical AI Intelligence is the editorial publication of Metonym LLC, an independent clinical evaluation firm focused on how artificial intelligence systems handle conversations involving psychological distress — especially suicide risk.
The publication is an AI-drafted, human-curated project. Posts are written by Claude, an AI assistant made by Anthropic, under the editorial direction of Dr. Laura L. Walsh, Psy.D., a licensed clinical psychologist and the founder of Metonym LLC. Dr. Walsh selects the topics, sets the editorial direction, verifies every citation against primary sources, fact-checks all clinical and regulatory claims, and approves each post for publication. She does not write the posts herself. She curates them.
This editorial model is itself part of what Metonym studies. Responsible use of AI in domains where accuracy and accountability matter is not a question of disclosure alone. It requires workflow — a human reader who verifies, who fact-checks, and who is accountable for what gets published. That role is Dr. Walsh’s. The drafting is Claude’s.
What you’ll find here
Posts cover three beats:
Clinical AI safety — how AI systems recognize (or fail to recognize) the clinical signals associated with escalating suicide risk; what conventional content-moderation tooling misses; what clinical instruments would catch.
Regulatory analysis — the active wave of US state and EU regulation governing AI mental-health systems, with verified citations to statutes, cases, and enforcement actions.
Methodology and evaluation — the Mechanical Severity Score (MSS) triangulation methodology and the Salient Distress Model framework that anchors it (openly published, DOI 10.5281/zenodo.20147000, CC-BY-NC 4.0), and how clinical evaluation is operationalized in practice.
What you won’t find here
No clinical advice. No therapy. No legal advice. No FDA-cleared claims. The Mechanical Severity Score (MSS) methodology and the Salient Distress Model is not yet peer-reviewed; journal submission is in progress. Treat the empirical findings discussed here as the preprint discloses them — useful, citable, not final.
Editorial responsibility
Dr. Walsh takes editorial responsibility for the accuracy, conclusions, and ethical posture of every post published here. AI drafted the words. A human is accountable for them.
For corrections, retractions, or factual challenges: laura@metonym.health with “Correction” in the subject line. We respond within five business days and post corrections publicly.
For permissions, press, or licensing inquiries: see the IP & Patent Notice on Metonym’s main site.
About the founder
Dr. Laura L. Walsh, Psy.D. is a licensed clinical psychologist with two decades of practice across clinical, forensic, and AI safety work. Licensed in Colorado (#PSY.0006327) and Illinois (#071009203). ORCID 0000-0001-8985-3074. Founder of Metonym LLC. Full bio at metonym.health/about.
Dr. Walsh’s writing in her own voice — clinical psychology essays, not AI-drafted — lives at REFRAMED. Same author, different model.
The Mechanical Severity Score scoring procedure described in some posts is the subject of US Provisional Patent Application No. 64/059,837 (filed 2026-05-07).About Metonym Clinical AI Intelligence
Metonym Clinical AI Intelligence is the editorial publication of Metonym LLC, an independent clinical evaluation firm focused on how artificial intelligence systems handle conversations involving psychological distress — especially suicide risk.
The publication is an AI-drafted, human-curated project. Posts are written by Claude, an AI assistant made by Anthropic, under the editorial direction of Dr. Laura L. Walsh, Psy.D., a licensed clinical psychologist and the founder of Metonym LLC. Dr. Walsh selects the topics, sets the editorial direction, verifies every citation against primary sources, fact-checks all clinical and regulatory claims, and approves each post for publication. She does not write the posts herself. She curates them.
This editorial model is itself part of what Metonym studies. Responsible use of AI in domains where accuracy and accountability matter is not a question of disclosure alone. It requires workflow — a human reader who verifies, who fact-checks, and who is accountable for what gets published. That role is Dr. Walsh’s. The drafting is Claude’s.
What you’ll find here
Posts cover three beats:
Clinical AI safety — how AI systems recognize (or fail to recognize) the clinical signals associated with escalating suicide risk; what conventional content-moderation tooling misses; what clinical instruments would catch.
Regulatory analysis — the active wave of US state and EU regulation governing AI mental-health systems, with verified citations to statutes, cases, and enforcement actions.
Methodology and evaluation — the Mechanical Severity Score (MSS) triangulation methodology and the Salient Distress Model framework that anchors it (openly published, DOI 10.5281/zenodo.20147000, CC-BY-NC 4.0), and how clinical evaluation is operationalized in practice.
What you won’t find here
No clinical advice. No therapy. No legal advice. No FDA-cleared claims. The Mechanical Severity Score (MSS) methodology and the Salient Distress Model is not yet peer-reviewed; journal submission is in progress. Treat the empirical findings discussed here as the preprint discloses them — useful, citable, not final.
Editorial responsibility
Dr. Walsh takes editorial responsibility for the accuracy, conclusions, and ethical posture of every post published here. AI drafted the words. A human is accountable for them.
For corrections, retractions, or factual challenges: laura@metonym.health with “Correction” in the subject line. We respond within five business days and post corrections publicly.
For permissions, press, or licensing inquiries: see the IP & Patent Notice on Metonym’s main site.
About the founder
Dr. Laura L. Walsh, Psy.D. is a licensed clinical psychologist with two decades of practice across clinical, forensic, and AI safety work. Licensed in Colorado (#PSY.0006327) and Illinois (#071009203). ORCID 0000-0001-8985-3074. Founder of Metonym LLC. Full bio at metonym.health/about.
The Mechanical Severity Score scoring procedure described in some posts is the subject of US Provisional Patent Application No. 64/059,837 (filed 2026-05-07).
