ChainIT Releases Provable Compliance Protocol for Agentic Commerce

ChainIT, a digital verification and compliance platform, has unveiled a new protocol aimed at enhancing transactional controls in the agentic commerce environment. The protocol, dubbed "Provable Compliance," is designed to address the complex issue of runtime compliance evidence and transaction-specific decision-making. This development is crucial as agentic commerce, which leverages AI agents to expedite and scale transactions, does not eliminate compliance requirements but rather intensifies the need for real-time assessments.

According to the press release, "Provable Compliance" is a technical white paper that accompanies "Provable Authority," another document ChainIT released last week. This new protocol focuses on a critical question: even when valid authority exists, do current compliance evidence and the organization's applicable policy permit this specific action to proceed now? The protocol is designed to evaluate compliance evidence at the moment of reliance, ensuring that decisions are current, attributable, and auditable.

The "Provable Compliance" protocol operates within ChainIT's Transaction Truth Architecture, a zero-trust framework for transaction control, execution, and evidence. It uses modular Evidence Assertion VDTs, Validated Data Tokens, to preserve the source class, provenance, observation time, confidence, permitted purpose, and privacy classification. A versioned Compliance Policy Profile selects the applicable requirements, and the ChainIT Business Rules Engine returns explicit, reason-coded predicates. The organization's Workflow Engine then produces a signed Compliance Decision VDT, which can include dispositions such as Allow, Allow with Controls, Step-Up, Review, Hold, Reject, Prohibit, or Stale/Unknown.

Jeremy Blackburn, CEO of ChainIT, emphasized the importance of this protocol: "Last week, we addressed who authorized the AI agent and what the agent was permitted to do. Provable Compliance addresses the next question: even when authority exists, do current compliance evidence and the organization's policy permit this specific action right now?" He noted that agentic commerce does not eliminate compliance requirements; it compresses the time available to evaluate them. AI agents should not operate based on a check performed weeks or months earlier, and a favorable result should never become standing authority to move money.

Russell Lessard, Chief Compliance Officer at ChainIT, explained that compliance is not a permanent badge that follows a person, business, or AI agent everywhere. He said, "Sanctions, AML monitoring, PEP risk, identity assurance, licensing, beneficial ownership, and payment-transparency requirements all operate differently. A proposed action may be allowed, allowed with controls, held for review, or prohibited depending on the current compliance evidence, jurisdiction, and policy."

The framework is intended for financial institutions, payment companies, stablecoin issuers, marketplaces, technology providers, and enterprises evaluating activity initiated by people, groups, organizations, organization-controlled workflows, and AI agents. It does not replace a regulated institution's AML/CFT, sanctions, or other compliance program and does not create a universal legal determination. Instead, it provides an evidence-and-decision layer that can be integrated with existing institutional policies, review processes, and execution controls.