When Depression Meets Digital Gatekeepers: Clinicians Sound Alarm on AI Mental Health Triage
Frontline workers at major healthcare systems report that automated intake tools are missing suicide risk, delaying urgent care, and routing severely ill patients to meditation apps.
Mental health clinicians at two major U.S. healthcare organizations are raising concerns that AI-powered triage systems are creating gaps in psychiatric care, with automated tools allegedly missing suicide attempts, routing severely suicidal patients to inappropriate care levels, and imposing navigation burdens on patients whose depression symptoms make such tasks difficult.
The warnings come from workers at Kaiser Permanente and Rogers Behavioral Health, where algorithmic intake and triage systems have been deployed to manage patient flow. Their accounts describe what they characterize as a troubling pattern: technology designed to improve efficiency may be undermining clinical judgment.
Staffing Changes and Digital Tools
At Kaiser Permanente's Walnut Creek Medical Center, the triage team has reportedly shrunk from nine clinicians to three over the past three years, with automated and algorithmic tools absorbing much of the workload. Harimandir Khalsa, a triage clinician at the facility, reports that approximately one-third of her roughly 24 daily calls now involve patients struggling to access appropriate care after being routed through digital pathways.
Khalsa describes delays of "sometimes weeks" with the new system and reports that indicators of self-harm and prior suicide attempts have been missed. Patients are frequently directed to third-party apps including Calm for guided meditation and Meru Health, a 12-week program offering up to four telehealth sessions.
Ilana Marcucci-Morris, a therapist at Kaiser Oakland, reports seeing patients who completed these 12-week digital programs in worse condition than when they started—an outcome that suggests to her that algorithmic routing may be substituting convenience for clinical appropriateness.
The Navigation Challenge
The clinicians identify what they see as a fundamental design flaw: patients whose depression includes lack of motivation—a core symptom of the condition—are being asked to navigate complex digital intake processes that require sustained effort and engagement. This creates what might be termed a "motivation paradox," where those most in need of immediate clinical intervention face the highest barriers to accessing it.
Melissa Stevens, a clinical psychologist in Kaiser Northern California's chronic-pain program, has observed this dynamic as automated systems impose navigation burdens on patients she believes are least equipped to manage them.
Rogers Behavioral Health Implementation
Rogers Behavioral Health, a Wisconsin-based organization operating across 10 states, partnered with AI vendor Limbic in late 2024 to deploy an AI intake chatbot and voice agent across its facilities. A Limbic case study from August 2025 claims the system provides "real-time routing and prioritization based on AI-generated clinical intelligence."
Frontline staff report concerns about the system's performance. Erin Quinlan, a behavioral specialist at Rogers' Madison, Wisconsin location, estimates that at least one of every eight patients in group classes has been placed at the wrong care level. Kate Zolandz, a former therapist who worked at Rogers West Allis, Wisconsin for four years, corroborates these concerns about inappropriate placements.
Regulatory and Labor Context
The controversy emerges as California considers legislation to establish AI safeguards in medical settings. Union representatives are pursuing collective bargaining goals that include transparency requirements and blocking AI from making triage decisions without clinical oversight.
The tension reflects a broader question facing healthcare: whether operational efficiency and cost reduction can be pursued through automation without compromising the clinical judgment that complex psychiatric cases require—particularly when the consequences of algorithmic error may include delayed care for suicidal patients and clinical deterioration during waits for appropriate treatment.
Sources
Companies in this article
LimbicNew York, NY, USA · Founded 2018Limbic is a clinical mental health AI company providing AI-powered support to increase access, speed up care, and improve outcomes for patients.View company profile →
Meru HealthMeru Health provides a therapist-guided digital mental health program that combines behavioral therapy, mindfulness practices, and biofeedback technology. The 12-week program treats depression, anxiety, stress, and burnout through personalized care delivered via mobile app.View company profile →Reporting & accountability
How we maintain accuracy
Health AI Central separates reported facts, analysis, and public-source data. Readers can review our standards, request a correction, or consult the relevant research methodology before relying on a claim.