Short answer
ATCE, the Adaptive Trust, Compliance and Execution Engine, differs from other checks mainly in timing and authority. It rules inside consensus before a transfer exists, and no transfer on the live chain moves value outside an ATCE ruling. Other approaches check after settlement, at one institution, on a message, or on an address. ATCE has costs, and the live chain has one operator and no independent audit.
What is being compared, and how?
This page compares ATCE with six categories of control, not products. No vendor, bank or product is named and no category is called bad.
A ruling is ATCE's decision on one act, made by every validator from the same inputs. It returns allow or deny, a numbered reason code, and one SHA-256 commitment binding the profile, the context and the outcome. A sovereign site is one jurisdiction's gated section of the network, with its own law.
Status, 2026-11-02: the chain jil-mainnet-2 is live, with ten validators operated by JIL Sovereign Technologies, Inc. on one hosting provider. ATCE enforcement is on for every transfer on it. The attester, the reviewer bridge and the agentic review are live. The record writer is built and not deployed.
How do the six categories compare with ATCE?
The table sets ATCE beside each category. Cells describe the typical shape of a category, and individual systems vary.
| Approach | When it runs | Who can refuse | What is recorded | Who keeps it | What a regulator can inspect | Can it question its own rules? |
|---|---|---|---|---|---|---|
| ATCE ruling | Before the act exists, in consensus | Every validator, from the same inputs. A review can only deny or hold an act the rules passed | Reason code and commitment on the chain. Full record designed for a write-once store | Commitment on the chain. Record body held by JIL, designed, allocated for each site at licensing | Recompute the commitment and compare it with the chain reference | Calls an agentic review that can hold an act. Live on jil-mainnet-2 |
| Compliance after settlement | After value has moved | A compliance team, by recall or enforcement | Reports and case files | The supervised firm | Reports the firm prepares | Built to test outcomes against rules; reviewers may question rules in a later cycle |
| Token allowlist | At transfer, in one token's code | The contract, by its administrator's list | Transfer events and list state | The chain the token sits on | The list and events for that token | Built to apply the list as written |
| Bank monitoring | During or after processing, at one institution | The institution | Alerts, case notes, filings | The institution | The institution's files, on request | Built to flag unusual behavior, with analysts judging |
| Travel-rule messaging | Alongside a transfer | Each institution, on the message | The messages | Each institution | Each institution's message log | Built to carry data; each party judges |
| Address screening | Before or after a transfer | The screening party | Screening results, scores | The screening party | The screening party's records | Built to score addresses against lists and behavior |
| Multilateral-style rulebook | By membership, then at each member | The governing body and members | What the rulebook requires | The members | Members' records and reporting | Built to let a governing body interpret and amend |
Can ATCE question its own rules?
ATCE calls an agentic review before its final ruling, and that review can hold an act that satisfies every rule yet looks wrong. The review is hold only: it can deny or hold an act the rules passed and it cannot release an act the rules denied. A person decides after a hold. ATCE's outcome is allow or deny, and public copy calls a deny that waits for a person a hold (reason 10, review denied, or reason 9, review missing). Read the agentic review page.
Status, 2026-11-02: the agentic review is live on jil-mainnet-2. The last column says what each other category is built to do.
How does ATCE differ from compliance applied after settlement?
ATCE decides before value moves, while after-settlement compliance reconstructs whether it should have. After-settlement review catches patterns no single transfer shows, and its remedy is recall, reversal or enforcement.
ATCE places the control at the only point where refusing is still possible. A refused act reverts with a reason code and a commitment, and a missing input denies. A site can run both.
How does ATCE differ from an allowlist inside a token contract?
An allowlist lives in one token's code and governs that token, while ATCE is a ledger-level ruling that applies to every ruled path. An allowlist is simple, cheap and often sufficient for one instrument on a chain that otherwise has no rules.
ATCE reads both parties' identity records, freeze flags and zones, plus the site profile, and it covers native transfers, token transfers, currency transfers and corridor releases. In ATCE, identity facts come from records an authorized attester signed. The trade is generality: the JIL ledger has no general-purpose virtual machine.
How does ATCE differ from bank transaction monitoring?
Bank monitoring is one institution watching its own customers across its own books, while ATCE rules on each act by the same published policy on every validator. Monitoring brings customer history and investigator judgment, which ATCE does not replicate.
ATCE does not replace a compliance officer. It enforces the rules the site operator publishes, and local law still governs. An ATCE ruling is deterministic and reached again identically by anyone replaying the block. The two address different failures. Monitoring finds what looks wrong over time. A ruling stops an act that the published policy does not allow.
How does ATCE differ from travel-rule messaging between institutions?
Travel-rule messaging carries originator and beneficiary information between institutions, while an ATCE ruling uses identity facts to decide whether the transfer may settle. Each party decides alone what to do with a message, and the message does not itself stop a transfer.
In ATCE the identity level, jurisdiction, risk score and sanctions flag come from on-chain records an attester signed, and a missing, revoked or expired record on either side denies. When value crosses between sites, the sending site's policy is enforced on departure and the receiving site's on arrival. Messaging works between institutions that share no ledger, which ATCE does not.
How does ATCE differ from public-chain address screening?
Address screening scores an address against lists and behavior, while ATCE rules on the act using attested records for both parties. Screening suits open chains where addresses carry no identity.
ATCE assumes the opposite arrangement: parties hold identity records, the policy is published, and the ruling is part of the application state. Screening gives a risk opinion held by the screening party. An ATCE ruling is a consensus result with a commitment. Screening reaches addresses ATCE never sees, because ATCE rules only on this ledger.
How does ATCE differ from a licensed multilateral-style rulebook?
A multilateral-style rulebook binds members by agreement and licensing, while ATCE binds transfers by code in consensus. JIL's public copy describes its structure, in which sites keep their own law and share one institution and corridors, as an analogy of structure only. JIL is not affiliated with any multilateral institution.
A rulebook relies on members to apply it and on a governing body to enforce it. ATCE makes the published policy of each site the policy that actually executes, and the site operator, not JIL, writes that policy. A governing body can grant exceptions and adapt without a software change. ATCE changes its profile only through a governed configuration transaction.
What does ATCE cost?
ATCE costs a licensed core operated by JIL, a ledger without general contracts, extra steps before a transfer settles, one operator on the live chain, and no independent audit yet.
- A licensed core operated by JIL. ATCE is patent pending and part of the Layer-1 that JIL Sovereign Technologies, Inc. operates, and a site cannot swap in a different ruling engine.
- A ledger that does not run general contracts. The Layer-1 has no general-purpose virtual machine. Instruments are issued through the currency platform rather than through arbitrary contract code.
- Refusal latency. A ruling needs identity records on the chain before the act, and a signed review result for the act. Without an attested identity record an act is denied (reason 8), and without a review result it is denied (reason 9), so if the reviewer service is down, off or unsure the act is denied. That is fail-closed by design, so an outage of a supporting service stops value movement. No latency figure is established.
- A single operator. All ten validators on jil-mainnet-2 are operated by JIL Sovereign Technologies, Inc. on one hosting provider. Validator agreement does not protect against a malicious operator.
- No independent audit yet. The consensus engine is self-written, has had no independent audit, and there are no external verifiers.
Status, 2026-11-02: the attester, the reviewer bridge and the agentic review are live. The record writer is built and not deployed, and the production record store is designed. JIL commits to keep each record for 20 years or more, and storage is allocated for each site when its license is signed.
When is ATCE the wrong choice?
ATCE is the wrong choice when the work needs open contract logic, a ledger operated by many independent parties today, an audited system today, or controls that reach beyond this ledger. In those cases another category fits better, and several can run alongside ATCE.
- You need arbitrary programmable contracts on a general-purpose chain. ATCE belongs to a ledger built without them.
- You need independent validators run by unrelated parties now. The live chain has one operator.
- Your procurement requires an independent audit of the consensus code today. None has been done.
- You supervise activity on other networks or in other institutions' books. Monitoring, messaging and screening reach there, and ATCE rules only on its own ledger.
The categories can be combined: a site using ATCE can still use monitoring, messaging and screening.
Related pages
Questions people ask about these comparisons
Can an allowlist do what ATCE does?
For one token on a chain that has no rules of its own, an allowlist can be enough. ATCE rules every ruled path on the ledger, reads both parties, and applies the site profile. It does so by ledger design, and that design means the ledger runs no general contracts.
Is ATCE like a multilateral institution's rulebook?
The structure has a likeness: sites keep their own law and share one institution and corridors. JIL describes this as an analogy of structure only and is not affiliated with any multilateral institution. ATCE enforces policy in code, while a rulebook relies on members and a governing body.
What are the main costs of choosing ATCE?
A licensed core operated by JIL, a ledger with no general contracts, a fail-closed dependence on supporting services, one operator for all ten validators, and no independent audit yet.
When would a team not choose ATCE?
When it needs general-purpose contracts, validators run by unrelated parties today, an audited consensus engine today, or controls over other networks and institutions. Those needs point to other categories, and some can sit beside ATCE.