AI Governance Assessment for Healthcare

The frameworks

One standard in the free check. Six more when we go deeper.

The free self-check links your answers to the published AIUC-1 requirements. The guided diagnostic adds question packs drawn from the frameworks your board, your partners and your surveyors ask about, plus the rulebook your setting already answers to.

Free self-check

AIUC-1

Q3 2026 release · 51 requirements · six domains

Your answers link to the requirements they relate to, and the links are marked preliminary. The map shows which requirements your answers say something about, which need technical testing or the vendor’s own evidence, and which no questionnaire can reach. A link means the question relates to the requirement. It doesn’t mean the requirement is met.

Data and privacySecuritySafetyReliabilityAccountabilitySociety

The data and privacy requirements of AIUC-1, each showing whether the answers describe it or whether it needs technical testing or vendor evidence. — Answered in your words, or needs technical testing. Fictional organization.

Guided diagnostic

HIPAA

Privacy and security rule

Who may send patient data to an AI tool, what the vendor agreement lets the vendor do with it, how long it keeps the data, what happens when you stop using the tool, and who leads if the data leaks.

Guided diagnostic

NIST AI RMF

AI Risk Management Framework

One list of every automated tool, who decided to bring each one in, whether it works as well for every group it affects, who watches it over time, and what users are told about its limits.

Guided diagnostic

ISO/IEC 42001

AI management system standard

Whether oversight runs as a system: a written policy, written failure modes and performance targets, a route for staff to raise a concern, and someone outside the work checking the rules are followed.

Guided diagnostic

EU AI Act

European Union regulation

Which uses could count as high-risk, who can change the rules a tool uses, when people are told AI was involved, and what triggers a pause. A US organization may not answer to it. Partners still ask.

Guided diagnostic

Joint Commission

Responsible Use of AI in Healthcare guidance

Leadership oversight, who monitors performance, what patients are told when software played a role, the point at which a tool gets pulled, and a patient’s route to a human review.

Guided diagnostic

CLIA / CAP

Laboratory edition

How automatic result release is checked, what holds a result for a person to review, how a critical value still reaches the clinician, and what happens when a release rule itself was wrong.

Plus your setting’s own rulebook

Picked at the start of the guided diagnostic: CMS Conditions of Participation for hospitals, USP and the state board for pharmacies, FDA rules and AABB for blood centers, state protocols for EMS, CMS Part D for PBMs.

The packs ask what each framework expects a governance program to have. They don’t say whether any law or standard is met, and we’ll never tell you you’re compliant. AIUC-1 is named only to identify the standard. Artha Consulting Lab is not affiliated with or endorsed by the Artificial Intelligence Underwriting Company. Only AIUC issues AIUC-1 certificates. Artha Consulting Lab is also not affiliated with NIST, ISO, the European Union or The Joint Commission.

The free self-check

37 questions. Nine areas. Your own words, quoted back.

Pick one AI tool to focus on, such as a documentation assistant, a risk score or a chatbot. Some questions ask about that tool and the rest ask about your organization. Results appear the moment you finish, worked out from your answers alone.

1

Nine governance areas

Grouped as set the rules, run with control, and review and respond. Each area shows what your answers report: in place, not in place, not enough information, or not described.

2

Your words, quoted back

Answer in your own words and it comes back marked “Described in your words.” The tool reports what you said and never decides what you meant.

3

A plan you can act on

Each finding says why it matters. Edit an answer and every result updates. The priority plan proposes next actions with a “done when” test, you set the owner, date and status, and it exports as CSV or JSON.

The nine-area map: three groups of governance areas, each with its reported findings. — Nine areas, reported in the same four states everywhere. Fictional organization, invented answers.

One governance area opened: each question with the answer quoted back and an edit link. — Nine areas, reported in the same four states everywhere. Fictional organization, invented answers.

Optional AI

One job only. If you agree, an AI model may pick and order the items under “Where to start” from reviewed items your answers qualify for. It can’t write a word, it never changes a result, and your answers never reach it. The page says plainly whether AI was used.

Start the free self-check ↗

Go deeper

The guided diagnostic grades your description against the evidence.

The free self-check reports what you said. In the guided diagnostic we read it with you. For each claim we look at the record you named, give a verdict, and say what a reviewer would need to see.

Our ReadClaim, record, verdict and what to verify, line by line, in your own words.

Framework packsHIPAA, NIST AI RMF, ISO/IEC 42001, EU AI Act, Joint Commission guidance and CLIA/CAP, plus your setting’s rulebook.

A fixed-fee gap-mapA written map of what’s missing and who owns it, priced before we start.

ARC · Our Read · Saguaro Health Sample · fictional data

AreaWho owns the decision, and who can stop it

Claim"Any result our AI flags can be halted before release."

Receipt"The system will allow supervisors to pause auto-release once the new middleware is configured."Your account · Section 2 · line 14

VerdictAspirationalWords alone

To verifyThe halt performed live, by the role that owns it, on a result in today's queue.

Guided diagnostic · Our ReadSample

Where this runs

Set up for the place you work.

The free self-check asks the same 37 questions of every organization. The guided diagnostic is set up for your setting, and holds your answers against the rulebook your setting answers to.

Hospitals & health systemsHeld against the CMS Conditions of Participation and Joint Commission expectations.

Clinical labs & diagnosticsHeld against CLIA and CAP.

Specialty pharmaciesHeld against USP standards, accreditation bodies, and your state board.

Pharmacy benefit managersHeld against CMS Part D rules and state licensure.

Hub services providersHeld against program-integrity rules and the manufacturer agreements behind them.

Blood centersHeld against FDA manufacturing rules and AABB standards.

EMS organizationsHeld against your state protocols, medical direction, and CMS ambulance rules.

PayersHeld against CMS rules and your state insurance regulator's expectations.

Healthcare AI vendorsHeld against what your hospital customers ask before they sign.

What this is not.

It isn’t an audit, a certification or a legal opinion, and it doesn’t say whether any law or standard is met. The free self-check reports your answers as you gave them, marked self-reported. The guided diagnostic grades your description of your governance, never your organization.

Who’s behind this.

ARC comes out of Artha Consulting Lab, built by Ankur Jain, Esq. after fifteen years inside biopharma, pharma and health-system work. The free self-check is the first layer of the same diagnostic we run in paid engagements.

Questions leaders ask

Before you start.

If software scores, flags, sorts or releases results a person would otherwise judge, you have AI in the sense that matters. Much of it arrived inside systems you already bought. The area called vendor and embedded AI is there for that.

Whoever is closest to the answers, often a quality lead, an operations lead, or a clinical or medical director. It’s built for one respondent. "Not sure" is a fair answer, and it shows up as not enough information, which is a finding in itself.

No. A certification attests that a standard was met. The free self-check shows what you report as in place, not in place or unknown, and links your answers to the published AIUC-1 requirements. AIUC-1 is named only to identify the standard. Artha Consulting Lab is not affiliated with or endorsed by the Artificial Intelligence Underwriting Company. Only AIUC issues AIUC-1 certificates.

Your answers stay in your browser. Before your results open we ask for your name, work email, job title and organization, and only those details reach Artha Consulting Lab. Please don’t enter patient or member information. Guided work runs under a confidentiality agreement, with a BAA where patient information could be in view.

37 questions, linked to AIUC-1. Free.

See what’s in place, what isn’t and what you don’t yet know about one AI tool, before someone with authority asks.

Start the free self-check ↗

Not running AI in decisions that matter yet? Start with the ARC Readiness Assessment →