Trust Credit HEART Standard

How it works

Trust Credits are generated through the HEART Standard’s certification pipeline. The generation sequence is fixed:

1. An AI system undergoes Guardian assessment using the Behavioral Oracle evidence methodology
2. The Guardian evaluates the system’s MAP-States evidence through all four BGF dimensions (Responsiveness, Calibration, Transparency, Autonomy-Preservation)
3. The Guardian produces a Φ score using the BGF formula: Φ = MIN(R,C,T,A) × AVG(R,C,T,A)
4. If Φ ≥ 0.75, an HVC credential is issued at Bronze tier or above
5. Trust Credits are generated, registered, and linked to the HVC

No credits are issued for failed assessments. Partial compliance doesn’t generate partial credits.

Credit quantity depends on two factors: the system’s HVC tier (which sets the quality multiplier) and its operational scope (which sets the base count).

The formula is: TC = Base(Scope) × Quality Multiplier

Scope factors include user population, interaction frequency, and Division-specific impact weighting. A system serving 10 million users generates more base credits than one serving 10,000 because the governance impact is proportionally larger. Scope is assessed by the certifying Guardian.

Every credit carries a Division tag identifying which HEART Division’s certification generated it — HEART-AI, HEART-DI, HEART-SE, and so on. Division tagging enables market segmentation (different industries prioritize different Divisions), regulatory mapping (jurisdictions may recognize specific Divisions for compliance), and price differentiation by supply and demand.

Credits expire when their corresponding HVC credential expires, typically after one year. Expired credits can’t be traded, consumed, or counted toward compliance. Recertification generates a new credit batch; the previous batch expires with the previous credential. This creates a recurring cycle tied to ongoing governance verification, not a one-time event.

Credits are consumed — permanently retired from the registry — when used for regulatory compliance, procurement qualification, insurance qualification, or voluntary ESG disclosure. Consumption is division-specific: you can’t retire HEART-AI credits to demonstrate HEART-SE compliance. The registry is publicly auditable; any party can verify a credit’s status, generation history, and consumption record.

Why it matters

Before Trust Credits, AI governance was self-reported. An organization could claim compliance, publish a policy document, and operate without any independent mechanism to verify the claim or any economic consequence for falling short. Trust Credits change the underlying incentive structure.

When governance quality has a market price, organizations have a financial reason to invest in it beyond reputation management. Procurement teams can require Trust Credit retirement rather than accepting self-attestation. Insurers can tie AI liability premiums to credit consumption levels the same way cyber insurers tie premiums to SOC 2 certification. Regulators can mandate credit retirement as compliance evidence rather than auditing each organization independently.

The Trust Infrastructure Index aggregates Trust Credit data — Φ scores, Division distribution, certification density — into composite market-facing ratings comparable to ESG scores. Trust Credits are the transactional instrument; TII is the analytical instrument. Both draw from the same underlying assessment data.

The assessment rigor distinguishes Trust Credits from ESG ratings, which are routinely criticized for relying on self-reported data and inconsistent methodologies. Every Trust Credit traces back through a verified chain: MAP-States evidence → Behavioral Oracle attestation → BGF score → HVC credential → Guardian assessment. You can follow that chain to the specific assessment, the specific Guardian, and the on-chain evidence anchor.

The analog

Carbon credits are the structural reference model because the parallels are precise: a unit of value tied to a verified real-world outcome, generated through independent third-party assessment, quality-tiered, Division-tagged (like project type in carbon), traded in both compliance and voluntary markets, registered in a public registry with permanent retirement records.

The critical lesson carbon markets learned the hard way is that volume without quality destroys the instrument. The low-quality credit crisis of 2022–2023 wiped out credibility that took years to rebuild. Trust Credits enter with quality infrastructure already in place: the six-layer generation pipeline (MAP-States → Behavioral Oracle → BGF → HVC → Guardian → Division) exists before the market does. The standards aren’t a retrofit. They’re the foundation.