A regulatory-compliant AI and verification system for higher education under ESG-aligned constraints

Abstract

This paper introduces a formal system model for integrating artificial intelligence and distributed verification into
higher education under European regulatory constraints. The architecture consists of two interlinked components:
EduAI, a role-specific, multi-agent artificial intelligence framework for institutional operations; and EduDVS, a
decentralised verification infrastructure for regulatory audit, credential authentication, and tamper-evident recordkeeping. The model encodes legal instruments such as the GDPR, EU AI Act, EQF, ECTS, and ESG directives
as structural system constraints, with additional compatibility for financial frameworks including MiFID II, MiCA,
AMLD, DORA, and Swiss equivalents (FMIA, FinSA, FADP). EduAI agents are formalised by stakeholder class
and governed by a constrained optimisation function ensuring legally admissible outputs. EduDVS operates as a
DAG-based, permissionless ledger maintained by a federated educational consortium, supporting verifiable academic
tokens, programmable stablecoins, and audit-ready interactions. Results are presented as a theoretical framework for
compliant digital infrastructures, with direct applicability to cross-jurisdictional academic ecosystems. The model
provides a foundation for regulatory prototyping, governance simulation, and controlled empirical validation.