For Funders

The Heart AI Foundation is building forensic audit infrastructure for AI behavioral evidence. Funding accelerates the public-interest infrastructure that lets AI governance claims become reviewable, attestable, and independently assessed.

Why this matters now

AI governance is moving from principle statements to evidence requirements. Regulators, insurers, procurement teams, and civic institutions need to know whether AI governance controls actually operate in deployment, whether evidence can be preserved, and whether independent reviewers can evaluate that evidence.

The missing category is forensic infrastructure: chain of custody, evidence preservation, governance-state attestation, human adjudication, reliability metrics, and credentials that third parties can verify.

The first fundable public artifact is the AI Behavioral Evidence Review Toolkit: a lightweight review method that turns AI behavioral artifacts into bounded evidence packets before full certification infrastructure is required.

Current reality

AreaCurrent status
Legal statusRegistered in Oregon as a nonprofit public benefit corporation
Tax statusFederal 501(c)(3) tax-exempt recognition pending
GovernanceBoard seating and independent director review underway
Primary fundable buildAI Behavioral Evidence Review Toolkit
Infrastructure wedgeGTE reference implementation and open trust infrastructure hardening
Research baseHEART Standard v1.8, Foundation Charter v2.4, ABTF/TRACE, MAP-States, RCTA/BGF validation path
Near-term funding needBootstrap capital for Toolkit, Guardian pilot, validation studies, insurance/procurement mapping, and registry path
Commercial boundaryFoundation maintains standards and public-interest infrastructure; Dwell/EMPI House are adjacent commercial/reference implementations, not the Foundation’s charitable program

What HEART contributes

HEART is not another ethics pledge. It is a forensic operational layer for AI governance.

ContributionWhat it does
GTEProtects governance controls and produces attestable execution evidence
MAP-StatesMakes AI behavior observable through structured evidence frames
Behavioral OraclePreserves and attests behavioral evidence
RCTA/BGFMeasures governance quality across Recognition, Calibration, Transparency, and Accountability
GuardiansCreates an independent professional review class
HVCConverts assessment into a market-legible credential
ABTF/TRACESupports investigative forensics when AI behavioral harm must be reviewed after the fact

What has already been built

Fundable milestones

AI Behavioral Evidence Review Toolkit

Build the first lightweight public entry point for civic, academic, journalism, compliance, and policy users who need forensic-informed AI behavioral review before full certification infrastructure exists. The Toolkit should produce templates, review forms, classification guidance, example packets, and training material.

Read the Toolkit page

GTE hardening

Harden the open trust infrastructure so deployers can protect governance controls for HEART and non-HEART frameworks. This is the adoption wedge.

Guardian pilot cohort

Train the first Guardian cohort, test assessment procedures, document reliability, and produce the first public case studies.

RCTA/BGF validation

Run reviewer-reliability, measurement-mode, and mechanistic correlation studies that make the scoring model funder-, regulator-, and insurer-legible.

Insurance and procurement mapping

Translate HVC tiers and continuous attestation into underwriting and procurement criteria that third parties can actually use.

Public registry and verification

Build the registry path for HVC credential verification, status visibility, revocation, and public accountability.

Heart City readiness

Prepare the municipal-scale deployment pathway for cities that need forensic audit infrastructure across civic AI systems.

Why the Foundation is the right vehicle

The infrastructure has to be public-interest, open enough to inspect, and protected against capture by the entities it evaluates. The Foundation exists to hold that boundary: open standards, independent Guardian practice, evidence discipline, anti-capture governance, and a clear separation between public infrastructure and commercial implementations.

Boundary with commercial implementations

Dwell, EMPI House, HeartCore Ventures, and related commercial/reference implementations are not the Foundation’s charitable program. They are adjacent deployments in the broader ecosystem. The Foundation’s role is to maintain standards, evidence methods, professional certification pathways, and public-interest governance infrastructure. Any related commercial entity can interact with the Foundation only as an external deployer, open-infrastructure user, or properly governed research-data contributor under the same terms available to similarly situated third parties.