standards and development targets · rules as code / RegTech / SDG
The register's unit of work and the unit of SDG Target 10.c are the same unit: the corridor.
This page says two things. Where the register sits in the rules-as-code and RegTech traditions, and which development targets its data can actually speak to. The second has a hard ceiling on how it is done: alignment is stated as subject-matter overlap, never as endorsement, and every row is required to write down what it does not claim.
the tradition it sits in
Making law machine-readable is not a new idea. It has several established lines, and this register stands deliberately on one of them and deliberately not on the others.
| Line of work | What it does | Where the register stands |
|---|---|---|
| Rules as code | Producing an authoritative machine-readable version of a rule alongside the rule itself, maintained by the body that owns it. | Method shared, standing different. The machine-readable version here is written by a third party, so it carries a sourcing obligation rather than authority: every cell must point back at the primary text so a reader can check it. |
| RegTech | Making it cheaper for a regulated firm to discharge its obligations: reporting, screening, monitoring. | One layer upstream. The register does not file or screen. It answers the prior question those tools assume is settled: which law actually binds on this corridor. |
| SupTech | Giving supervisors a data-shaped view of the market and of peer regimes. | Directly usable, with no claim to substitute. A cross-jurisdictional horizontal view is the register's shape. What it does not have is any non-public supervisory data. |
| Agent grounding | Giving an AI system that acts a record it can cite rather than paraphrase. | The reason the register exists in this shape, and the reason it ships as typed MCP tools rather than as a PDF. |
a deliberate limit
The register does not compile law into executable logic and does not claim to adjudicate whether a particular transaction complies. It encodes which rule binds on which edge, on the authority of which provision, and whether that provision is in force today. Conflating “which rule applies” with “is this transaction compliant” is the standard failure on this road, because the first is checkable and the second needs facts, intent, and an adviser who carries the liability.
Which development targets this data can speak to
The four rows below follow one structure, isomorphic with every other record in the register: the target, the mechanism in the register that corresponds to it, the records backing it, the strength of the claim, and what it does not claim. The last field is the most important one on this page.
Reduce remittance transaction costs, and eliminate corridors costing more than 5 per cent
Target. By 2030, reduce migrant remittance transaction costs to below 3 per cent and eliminate remittance corridors costing more than 5 per cent.
Mechanism in the register. The object of this target is the corridor, which is this dataset's unit. The G20 Roadmap for Enhancing Cross-Border Payments, coordinated by the Financial Stability Board, also pins its remittance-cost target to this indicator. Part of what sets a corridor's cost structure is regulatory antecedent: a corridor that needs an equivalence determination to enter is not priced like one where each end is separately authorizable. The register records that antecedent — which mechanism is in the way, on which provision, and the date it changes.
Backing records. 132 directed edges, 6 feasibility classes, 16 scheduled reclassifications. claim_class = tier1_legal.
What this does not claim. The register does not measure remittance cost, not by one basis point. It does not claim any corridor became cheaper because it exists, and it does not claim regulatory feasibility is the only or even the main determinant of cost. It claims one thing: the object this target names and the unit of this data are the same object.
Increase the availability of high-quality, timely and accessible data
Target. Increase the availability of high-quality, timely and reliable data, disaggregated along relevant dimensions.
Mechanism in the register. Data under CC-BY-4.0, code under Apache-2.0, a public schema, every release archived to Zenodo with a DOI, disaggregable by jurisdiction, dimension and directed edge. “Timely” here has a specific meaning rather than an adjectival one: every record carries the date it was read, and the build refuses to serve an uncommenced statute as current law.
Backing records. All 152 records and their schema; the 46-record citable subset; the published per-cell worklist.
What this does not claim. This is not official statistics, not part of any national statistical system, and not fit for use as a reporting source for SDG indicators. It is a third-party research dataset covering twelve jurisdictions, maintained by one person.
Significantly reduce illicit financial flows
Target. By 2030, significantly reduce illicit financial and arms flows, strengthen recovery and return of stolen assets, and combat organized crime.
Mechanism in the register. The AML / customer-due-diligence dimension and the records touching the FATF Travel Rule. Divergence between jurisdictions on these obligations is itself part of why flows route the way they do; writing each jurisdiction's stance on that dimension as comparable data is what makes the divergence visible at all.
Backing records. The per-jurisdiction records under the aml_kyc dimension and the Travel Rule entries.
What this does not claim. The register does not detect, measure or estimate illicit flows. It contains no transaction data, no chain analytics, and no entity-level information. It records obligations, not conduct, and makes no claim that recording an obligation reduces breaches of it.
Expand access to financial services
Target. Strengthen the capacity of domestic financial institutions to encourage and expand access to banking, insurance and financial services for all.
Mechanism in the register. The investor-gating dimension. A professional-investor gate and a transaction cap are legally different instruments with different consequences for access, and collapsing both into a single “restricted” cell erases precisely that difference. The register keeps them apart.
Backing records. The per-jurisdiction records under the investor-gating and retail-access dimensions.
What this does not claim. This causal chain is longer than the other three. The register does not measure financial access, does not claim stablecoins expand it, and takes no position on that question. It claims only that treating different restricting instruments as one thing is a misleading simplification in this debate.
the targets that are not on this page
SDG 1, SDG 5 and their neighbours are absent here not because they matter less, but because the causal chain from this data to those outcomes is too long to hold up under the register's own evidence discipline. An alignment that cannot be defended weakens the three that can, so it is left out.
For the same reason this page shows no UN emblem, no SDG colour wheel, and no goal icons. Use of those marks is governed by its own rules, and putting them on a third-party dataset invites “subject-matter overlap” to be read as “recognized by”. There is no affiliation here and no endorsement.
the mapping is data too
If this layer were only four paragraphs on a page it would be a coloured badge with more words. So it is written in the same structure as every other record, and ships in the next release the same way: queryable, diffable, arguable.
// analysis/sdg_mapping.json — one row per mapped target { "sdg_target": "10.c", "target_text": "…", "mechanism_in_register": "corridor_layer", "backing_records": ["corridors/*", "transitions/*"], "claim_class": "subject_matter_alignment", "what_this_does_not_claim": "no measurement of remittance cost", "as_of": "2026-08-20" }
claim_class takes the value subject_matter_alignment here, not tier1_legal. That is not a rhetorical distinction: it means citable_law() will never return these rows. The mapping can be read, cited and argued with, and it cannot enter the instrument a lawyer is shown.