Skip to content
Tyche Institute

Research domain

Governance

Technical systems do not govern themselves.

For Tyche, governance is the study of how authority, evidence, responsibility, and contestability are organised around technical systems. Cryptography can bind a record, but it cannot decide whether the underlying process was legitimate, whether authority was properly delegated, or whether a person had a meaningful route to challenge a decision. Those are governance questions.

Tyche therefore studies AI and digital trust as parts of public and organisational life: who may act, under what mandate, what evidence is retained, who can inspect it, and what happens when institutions or technologies change.

Current questions

Governance topics in the research programme

Public-sector AI accountability

What must be recorded when AI enters an administrative workflow, what a human reviewer actually saw and did, and how an affected person can contest the result.

Digital-state infrastructure

How identity wallets, PKI, X-Road-style exchange, trusted lists, and delegated machine action distribute institutional authority and shape administrative legibility.

Rules made operational

How regulation, standards, procurement, and organisational policy become testable controls—or fail to leave evidence that those controls ran.

Public evidence and institutional memory

How registers, lifecycle records, open data, preservation, and reproducible observatories make governance claims visible over time without turning records into verdicts.

Research posture

Evidence for scrutiny, not automated legitimacy

Tyche asks whether governance claims can be inspected. A statement such as “a human reviewed this decision” becomes more useful when the record shows where the system entered the workflow, what the reviewer saw, what action they took, and which contest or repair path remained available.

Verifiable records can support accountability, but they do not prove fairness, legality, truth, or democratic legitimacy. The institute does not provide legal advice, certify compliance, or treat technical verification as a substitute for institutional judgment.

Where to explore

Research, records, and live evidence streams

  • Research corpus →

    Working papers on public-sector oversight, digital identity, infrastructure governance, procurement, regulation, and agent accountability.

  • Public-sector AI oversight matrix →

    A comparative evidence matrix for how human oversight appears in public legal, regulatory, audit, and administrative records.

  • Observatories →

    Reproducible public-data streams covering procurement, registers, institutional boundaries, lifecycle change, and governance evidence.

  • Standards work →

    The formal venues where evidence formats, trust infrastructure, and AI governance controls become shared technical rules.