A National Initiative for Public AI Accountability

Why This Matters

Artificial intelligence is now infrastructure. It underwrites how Americans vote, learn, work, receive care, and encounter their government. Infrastructure of this consequence requires public evidence, independent measurement, and a durable record — the same things we already demand of bridges, banks, and clinical trials.

The Ethos AI Registry exists to provide that public record. It is independent, standards-aligned, and free to use. The case for supporting it is not a startup pitch. It is a case for civic infrastructure.

Partner with the RegistrySupport the missionRead the methodology

The public interest is unrepresented in the AI market

Vendors publish marketing. Regulators publish rules. Neither produces a persistent, citable record of which systems are deployed, how they perform, and what happens when they fail. That absence is not an oversight — it is a gap in national infrastructure.

A registry is the missing primitive

Just as the SEC's public filings, the FDA's adverse event system, and the NTSB's incident reports underpin trust in other sectors, a public AI registry produces the shared evidence base that policy, procurement, and civil society all depend on.

Where trustworthy AI matters

Nine domains, one shared evidence base

The domains below are not verticals to be sold into. They are areas of public life where the presence or absence of trustworthy AI produces measurable outcomes for Americans.

Democracy

What is at stake
Elections, civic information, and public discourse increasingly move through automated systems that shape what citizens see and believe.
Where trust breaks down
Opaque recommendation, moderation, and generative systems can amplify disinformation, suppress lawful speech, and erode confidence in institutions.
What the registry provides
Independent transparency scores and incident evidence give lawmakers, election officials, and journalists a shared factual record — not a vendor talking point.

Education

What is at stake
AI is now embedded in admissions screening, plagiarism detection, adaptive learning, and school operations from K–12 through higher education.
Where trust breaks down
Unaudited tools can misclassify students, penalize non-native English speakers, and expand achievement gaps behind a curtain of neutrality.
What the registry provides
A public registry lets districts, universities, and parents see which systems have been evaluated, how they scored, and where communities have reported harm.

Healthcare

What is at stake
Clinical triage, prior authorization, imaging, and care navigation are being automated at scale across payers and providers.
Where trust breaks down
Silent errors in diagnostic and eligibility models can delay care, deny coverage, and disproportionately harm patients who are already underserved.
What the registry provides
Standardized audit questionnaires and community incident reporting surface pediatric, disability, and language-access risks before they become mortality statistics.

Workforce

What is at stake
Hiring, scheduling, performance review, and termination decisions are increasingly mediated by automated systems.
Where trust breaks down
Résumé rankers and productivity monitors can encode historical bias, chill lawful worker activity, and shift risk onto workers with the least recourse.
What the registry provides
Employers, unions, and state labor agencies can consult a common evidence base rather than commissioning duplicative private audits behind NDAs.

Public Safety

What is at stake
Facial recognition, license-plate readers, gunshot detection, and predictive tools are deployed by agencies with lethal authority.
Where trust breaks down
Errors and misuse in these systems produce wrongful stops, arrests, and search-warrant applications — often with no route to appeal.
What the registry provides
The registry documents deployments, oversight practices, and reported incidents so procurement officers and civilian oversight boards can act on evidence.

National Security

What is at stake
Foreign-built and foreign-influenced AI systems are being integrated into critical infrastructure, logistics, and information environments.
Where trust breaks down
Supply-chain opacity, model provenance gaps, and unverified claims of alignment create strategic exposure that no single agency can map alone.
What the registry provides
A U.S.-based, standards-aligned public registry complements federal efforts by providing independent, citable evidence of who deployed what, where, and with what oversight.

Economic Opportunity

What is at stake
Credit, insurance, housing, and access to public benefits are gated by scoring and eligibility models.
Where trust breaks down
Disparate impact in these systems compounds across generations, quietly redlining ZIP codes and populations without a paper trail.
What the registry provides
Publishing audit scores and fairness disclosures creates a market signal: vendors that document their systems win trust; those that do not are visible as such.

Small Business

What is at stake
Small businesses buy AI they did not build — from point-of-sale fraud tools to marketing platforms — and inherit its risks.
Where trust breaks down
Without an accessible reference, a small operator cannot tell a well-governed vendor from a rebranded API wrapper, and bears the liability either way.
What the registry provides
A free public registry gives small businesses, cooperatives, and nonprofits the same due-diligence surface that Fortune 500 procurement teams already have.

Underserved Communities

What is at stake
The communities most affected by automated decisions are typically the last to be consulted about them.
Where trust breaks down
Language, disability, immigration status, and rural bandwidth gaps are routinely absent from vendor testing — and therefore invisible in vendor reporting.
What the registry provides
Community reporting is a first-class input to the registry, weighted alongside technical audits, so lived experience becomes citable evidence rather than anecdote.

Why independence matters

An independent public registry benefits everyone it does not belong to

A registry captured by any one interest — commercial, partisan, or agency — cannot serve the others. The Ethos AI Registry is built on four non-negotiable principles.

Independent

Operated by a 501(c)(3) public charity. No vendor, agency, or sponsor holds editorial control over audit outcomes or incident records.

Standards-aligned

Methodology maps to the NIST AI Risk Management Framework and complements ISO/IEC 42001 and emerging state and federal statutes.

Public by default

Registry entries, transparency scores, and community reports are free to read, cite, and reuse under an open license.

Evidence-first

Every claim on the registry links to a source: a questionnaire response, a public disclosure, or a verified community report.

How to support the initiative

Sponsorship, partnership, and collaboration

Financial support of the Ethos AI Registry is tax-deductible in the United States and disclosed publicly. Every partnership agreement preserves the registry's editorial independence.

Philanthropic Foundations

Sponsor a national public good

Underwrite audit capacity, community reporting infrastructure, and public education. Named sponsorship of methodology areas (governance, risk mapping, incident intake) is available.

Contact partnerships@ethosairegistry.org

Government Agencies

Interagency and jurisdictional partnerships

Federal, state, and municipal agencies can coordinate on procurement standards, incident reporting protocols, and reference implementations of the NIST AI Risk Management Framework.

Contact gov@ethosairegistry.org

Corporate Sponsors

Underwrite transparency, not advocacy

Corporate sponsors fund infrastructure — not editorial positions. Sponsorship is disclosed publicly; no sponsor receives preferential audit outcomes, embargoed access, or takedown rights.

Contact sponsors@ethosairegistry.org

Research Institutions

Research partnerships and data collaboration

University labs and think tanks can co-author methodology, contribute audits, and access de-identified incident and questionnaire data for peer-reviewed research under a data-use agreement.

Contact research@ethosairegistry.org

Universities

Course adoption and clinic collaboration

Public policy, law, computer science, and journalism programs can integrate the registry into coursework, clinics, and capstones. Teaching materials and audit rubrics are available at no cost.

Contact universities@ethosairegistry.org

Public–Private Consortia

Standards, benchmarks, and shared evidence

Join a working group aligning private-sector audit practice with NIST AI RMF, ISO/IEC 42001, and emerging state statutes. Consortium members shape published benchmarks.

Contact consortium@ethosairegistry.org

Grantmakers

Fund specific programs

Restricted grants are welcomed for community reporting, translation and language access, rural civic sandbox deployments, and public-interest audits of high-impact systems.

Contact grants@ethosairegistry.org

Common questions

Partnership FAQ

Eligibility, payment cadence, public credit, and the independence terms that protect the registry.

Have a question that is not answered here? Ask partnerships@ethosairegistry.org.

Fund the infrastructure the public interest requires

The Ethos AI Registry is a public good. It will exist at the scale its supporters make possible. If your institution shares the premise that trustworthy AI is national infrastructure, we would like to hear from you.

partnerships@ethosairegistry.orgMake a tax-deductible gift

The Ethos AI Registry is an initiative of the Algorithmic Accountability Project, a program of the TechGeekette Foundation, a US 501(c)(3) public charity.