DeFi Compliance After Unbundling: Enforcing Mandates Without Full Position Disclosure

TL;DR

• Unbundled finance separates transaction execution from the parties holding identity, risk, and authorization data, so you must define enforceable limits for every agent payment.

• The GENIUS Act requires monthly reserve reporting and technical capabilities to block transactions. Monthly attestations cannot verify whether an instantly settled payment followed its mandate.

• Pre-execution enforcement checks counterparty limits, transaction thresholds, and sanctions screening results before an agent-initiated transaction settles.

• Zero-knowledge receipts let counterparties and regulators verify that an approved mandate governed each transaction without exposing positions, policy thresholds, or counterparty data.

The verification gap in unbundled finance

Unbundled finance separates transaction execution from the information needed to authorize it. A traditional bank once kept customer records, risk decisions, authorization, and settlement within one institution. DeFi protocols, vaults, stablecoin issuers, and tokenized asset platforms distribute those functions across several parties. Each participant sees part of the transaction, but no participant necessarily sees the complete compliance state when value moves.

Continuous settlement makes fragmented oversight harder to operate. A stablecoin transfer or vault trade can settle onchain at any hour, and an AI agent can initiate another transaction without waiting for human review. Periodic reporting can establish reserve conditions at a reporting date or identify suspicious activity after settlement. It cannot establish that a particular transaction satisfied sanctions restrictions, counterparty limits, or spending thresholds before assets moved.

The GENIUS Act makes the timing problem concrete. Under the GENIUS Act framework, permitted payment stablecoin issuers must publish reserve composition monthly and have each report examined by an independent public accounting firm. The same law requires issuers to maintain technical capabilities and procedures to block transactions. Monthly attestation documents reserve backing periodically, while transaction blocking requires a control that can act when a payment occurs. Forthcoming regulations will determine more of the technical detail.

Fragmented data complicates that control. A sanctions provider may supply a screening verdict, while a wallet provider holds the agent’s authorization. The issuer may hold customer information and regulatory responsibility, while a protocol executes the transaction. Post-transaction monitoring can combine some of those records later, but onchain settlement may already be final by then.

You can regain control by putting an intermediary back in the transaction path and collecting sensitive data from every participant. Recentralization adds approval delays, creates concentrated data stores, and reduces the operational benefits of programmable settlement. Alternatively, you can require full disclosure of identities, positions, counterparties, and strategies. Full transparency gives reviewers more information, but it can expose confidential trading activity and deter institutional participation. Both workarounds treat fragmented oversight as a choice between centralized control and financial privacy.

What a mandate means for a vault, issuer, or protocol

A mandate defines the boundaries under which a vault, issuer, or protocol may move capital. The party granting authority sets objective conditions such as approved assets and counterparties, aggregate exposure limits, transaction thresholds, sanctions exclusions, and required approvals. A vault curator may publish these terms for allocators, while an issuer or platform may apply them to an AI agent that initiates payments.

An allocator might permit a vault to trade approved assets while keeping total exposure below a specified limit. The same mandate could require multiple signatures for withdrawals and prohibit transactions involving a sanctioned counterparty. The vault retains discretion over individual trades as long as each action stays within those boundaries.

Verification centers on mandate adherence because allocators often cannot inspect the vault’s positions or strategy. Full position disclosure could reveal trading methods, liquidity needs, and commercially sensitive relationships. An allocator instead needs evidence that each covered action satisfied the agreed conditions when the vault executed it.

A written mandate alone provides no evidence of compliance. The vault operator could describe strict controls while taking an off-mandate position between reporting periods. An allocator, lender, or regulator therefore needs a way to confirm that the mandate governed actual transactions without receiving the private data used to evaluate them.

For autonomous agents, the mandate also defines delegated authority in machine-readable terms. The payment platform can specify which counterparties an agent may use, how much it may spend, and which screening result must be present before execution. Compliance review then focuses on whether the agent stayed within that authority for each action.

Blocking violations before they settle

Pre-execution enforcement places a policy decision in the transaction path before an agent can commit funds. Post-transaction monitoring can flag suspicious activity and support investigations, but it cannot prevent an onchain transfer that has already reached final settlement. An inline control can reject the proposed action before the payment protocol submits it for execution.

Inherence converts the written policy into an enforcement gate that software can evaluate for each proposed action. When an AI agent initiates a payment through an agent payments protocol such as AP2, the gate checks the action against the applicable conditions and current compliance inputs. A passing action proceeds, while a failing action stops before execution.

Existing compliance providers can supply facts and risk decisions to the gate. For example, a sanctions screening service can return a verdict about a counterparty, and the gate can require an acceptable verdict before permitting payment. Identity credentials, transaction-risk scores, and approval records can serve similar roles. Inherence enforces the resulting policy decision rather than replacing those upstream services.

Placement determines what the control can accomplish. A monitoring tool observes an executed transaction and may open an alert afterward. A pre-execution gate evaluates the proposed transaction at the point where the agent still has the power to stop. That timing lets stablecoin issuers, payment platforms, and DeFi protocols apply compliance controls continuously to autonomous activity.

Publishers should state any latency or performance claim using Inherence’s canonical measured benchmarks and the relevant hardware context. Legal review should approve those claims before publication.

Proving compliance without disclosing positions

Zero-knowledge receipts let institutions verify compliance while keeping sensitive financial data private. Radical transparency would expose positions, strategies, thresholds, and counterparties. Recentralization would require one intermediary to collect that data and control execution, recreating the structure that decentralized rails were intended to avoid.

For each permitted action, a receipt confirms that the relevant mandate governed the transaction and that the action satisfied its constraints. An allocator could verify that a vault remained below an exposure limit without seeing its holdings. A payment counterparty could confirm that an agent passed transaction limits and sanctions requirements without learning the private threshold or counterparty identity.

Independent verification removes the operator from the trust path. Inherence produces a zero-knowledge receipt that an allocator, counterparty, auditor, regulator, protocol, or smart contract can validate directly. The verifier does not need access to the operator’s internal logs, a vendor portal, or the private inputs used during evaluation. Existing screening providers can supply sanctions or risk verdicts, while the receipt confirms that the required verdict governed execution.

Per-action evidence serves a different purpose than a monthly certified report. A monthly attestation describes compliance or reserves across a reporting period, but it cannot establish that a specific transaction satisfied its mandate when it executed. A receipt binds evidence to an individual action, while the pre-execution gate prevents actions that fail the policy from settling. Together, enforcement and verification provide continuous evidence without requiring continuous disclosure.

Where Sardine and ComplyAdvantage fit

Sardine supports fintech and crypto companies with AML, KYC, transaction monitoring, and sanctions screening. Its user-friendly interface and support experience make complex compliance operations easier to manage across customer onboarding and ongoing activity.

ComplyAdvantage provides broad financial crime risk intelligence across customer screening and transaction monitoring. Its automated tools help compliance functions identify sanctions exposure, suspicious behavior, and other risks across industries and regulatory environments.

Both vendors focus primarily on detecting, screening, and investigating risk around transaction activity. Those capabilities remain necessary, but monitoring can flag a violation after an autonomous agent has already initiated or settled a payment.

Inherence Labs adds a pre-execution policy layer for agentic commerce and stablecoin transactions. An agent payments protocol can use that layer to check counterparty restrictions, transaction thresholds, and sanctions rules before execution. The layer blocks actions that breach the mandate and produces privacy-preserving cryptographic receipts for permitted actions. Sardine or ComplyAdvantage can continue supporting identity checks, fraud prevention, risk intelligence, and investigations while Inherence enforces transaction-specific mandates before settlement.

What this means for stablecoin issuers and DeFi infrastructure providers

Pre-execution enforcement puts a compliance decision directly in the transaction path. A stablecoin issuer or DeFi platform can require an AI agent’s payment to pass defined controls before settlement. Identity services, sanctions screening, and transaction-risk tools remain part of the compliance stack, and their verdicts can feed the enforcement decision.

Verifiable receipts give compliance teams evidence for each covered transaction without requiring access to private positions, strategies, or counterparty data. Counterparties and regulators can independently verify that the applicable policy governed an action. Engineering leaders must decide where enforcement sits, which data sources supply policy inputs, and how their platform handles failed checks.

Trust can travel with the transaction instead of being reconstructed later through logs, disclosures, and investigations. Compliance and engineering leaders evaluating this layer can review Inherence.

FAQ

• How does Inherence differ from existing AML and sanctions screening tools? AML and sanctions tools identify customer, counterparty, and transaction risks. Inherence uses those screening verdicts within a mandate and blocks agent-initiated payments that fail its rules before execution. You can keep existing screening and investigation tools while adding transaction-time enforcement.

• Does Inherence replace GENIUS Act reserve attestations? Reserve attestations verify an issuer’s reserve composition and outstanding stablecoin supply at a required reporting interval. Inherence does not replace the GENIUS Act’s monthly disclosure and examination obligations. Issuers can use per-action enforcement alongside periodic reserve reporting to address separate compliance duties.

• What data can counterparties or regulators see in a receipt? A zero-knowledge receipt confirms that a covered action satisfied the governing mandate without revealing its private inputs. Inherence receipts can conceal amounts, thresholds, positions, strategies, and counterparty data while remaining independently verifiable. A verifier learns whether the action complied without receiving the underlying confidential records.

• Who defines the mandate that Inherence enforces? The organization delegating authority defines the written limits that govern an agent’s actions. Inherence converts those limits into controls covering areas such as transaction thresholds, permitted counterparties, sanctions requirements, and approval conditions. Compliance leaders retain policy ownership, while engineering leaders can apply the approved mandate consistently to each covered transaction.