# Technology: AI with evidence, structure and control

> The technology behind NormLogic: regulatory knowledge graphs, formal rule representations, domain-specific languages, provenance and temporal modelling, built on open standards.

Canonical: https://sireto.com/technology

## Approach

We use AI where probabilistic reasoning is useful, and formal systems where determinism, validation and auditability are required. AI extracts and assists. Structured models validate. Formal rules execute. Evidence provides source-level traceability.

## Six building blocks

- Regulatory knowledge graphs: requirements, concepts, actors, conditions, exceptions, dependencies and sources as connected knowledge.
- Formal rule representations: obligations, permissions, prohibitions, conditions and consequences expressed explicitly for deterministic evaluation.
- Domain-specific languages: regulatory semantics translated into representations software can validate, test and execute.
- Provenance: traceability from every assertion and decision back to the source provision, in line with W3C PROV.
- Temporal modelling: when a rule applies, when it changes and which version was in force at a point in time.
- Semantic interoperability: regulatory knowledge connected with open standards and existing enterprise and public-sector systems.

## Standards and interoperability

LegalRuleML (OASIS, legal rules); Akoma Ntoso (OASIS, legal documents); W3C PROV (provenance); prEN 18286 and emerging European work on machine-readable regulatory information; knowledge graphs and regulatory ontologies; formal rules and domain-specific languages. NormLogic enables interoperable representations of regulatory requirements that can be exchanged, validated and executed across systems and organisations.

## AI governance

Agents operating against explicit structured regulatory models linked to authoritative sources can ask what they are allowed or required to do, which conditions apply, which rule takes precedence, what evidence supports a conclusion and which source provision created an obligation.

## Auditable AI

Chain of evidence: Source → Interpretation → Rule → Reasoning → Decision. Conclusions can be inspected, challenged and reproduced.

## Regulatory digital twins

A regulatory digital twin is a continuously maintained computational representation of a regulatory environment, connecting legislation, regulatory guidance, standards, organisational policies, domain concepts, operational systems, compliance evidence and software agents. When a rule changes, its downstream impact becomes discoverable.