AI is making decisions about patients. AVAAS certifies the AI, not the patient data.

Clinical decision support, triage, diagnostic assistance, and coverage decisions increasingly run through AI. When that AI affects a patient, the stakes are health, not just liability. AVAAS healthcare certification is in development. When it launches, it will certify the behavior and governance of the system without requiring access to protected health information.

California AI denial limitGenerative-AI disclosureFDA considerationsNo PHI architectureSOC 2

Where AI acts on a person

The decision point in healthcare

In healthcare, the AI acts when it supports a diagnosis, triages a patient, or recommends approval or denial of coverage. Those decisions reach the patient directly.

When a model influences care or coverage, the decision has to be defensible to clinicians, payers, and the patient, and California now limits what AI can decide alone.

What keeps you exposed

What will keep providers exposed

Coverage denials

AI cannot be the sole basis for a denial

California now restricts AI from being the sole basis for denying a medical claim. Coverage models need human accountability and evidence.

Patient disclosure

Generative AI in patient communication

California requires disclosure when generative AI is used in patient communications. Undisclosed use is a compliance and trust risk.

PHI and trust

You cannot hand a patient data to a third party

AVAAS is being designed to certify the model, governance, and artifacts without taking PHI, so business-associate status never attaches.

This is already happening

California now restricts AI from being the sole basis for denying a medical claim and requires disclosure of generative AI in patient communications, while buyers ask for SOC 2 rather than a HIPAA certification that does not exist.

California SB 1120 · California AB 3030 · SOC 2

How AVAAS adds value

What AVAAS healthcare certification will provide

Does the model behave safely at the point of care?

Five structurally independent validators will test the system for the failure modes that matter in clinical and coverage settings using causal attribution.

Can clinical and payer governance see the evidence?

AVAAS will produce documented, third-party evidence suited to clinical governance, payer scrutiny, and patient trust.

Can you certify without exposing patient data?

A no-PHI architecture certifies the model, governance, and artifacts, returning aggregates rather than patient data.

AVAAS certifies conformity to a published standard. It is not a substitute for FDA clearance, and it is not legal compliance advice.

Related AVAAS coverage: California Healthcare · California ADMT. Or run the free Regulation Checker to see what applies to you.

Healthcare certification is coming soon.

Or email [email protected]