Three services. Each one fixed in scope.
Every engagement answers one question: what does a cyber or AI failure do to the patient? Each is scoped in writing before it starts, and none requires access to patient data.
Clinical AI Harm Review From $12,000
For digital health and AI vendors. An independent review of your ambient scribe, patient-facing chatbot or prior-authorization AI. I test it against real clinical workflows and score every finding by what it would do to a patient, mapped to the OWASP LLM Top 10 and NIST AI RMF. You get a Bedside Impact Report written for hospital security reviewers and AI governance committees.
The 3 A.M. Downtime Drill From $7,500
For hospitals and health systems. A clinical ransomware exercise for ED leaders, nursing and emergency management, run virtually. It covers what really happens when the EHR and alarm notifications fail at night: paper triage, unheard alarms, and a regional diversion cascade that sends unannounced ambulances to your door. You get an after-action report with owners and dates.
Fractional Clinical Safety Advisor From $3,500/month
For AI and health tech teams selling into hospitals. A clinical voice on your security, product and customer calls each month: hospital security reviews, AI governance questions, clinical risk input on releases, and security training your clinical users will actually absorb. Three-month minimum.
Medical device makers Through partners
The PSTM Clinical Harm Assessment adds patient-consequence scoring to your premarket cybersecurity threat model. It's delivered with regulatory and engineering partners. Ask about partnering.
Speaking & workshops By request
Talks and workshops on clinical AI risk and what cyber failures do at the bedside, from a 45-minute session to a half-day workshop. Email a request.