Live research databases - last synced March 2026

The empirical foundation
for court AI governance.

Brandeis TRP is not built on theory. It is built on 1,415 records across five live research databases tracking AI adoption, procurement, controversy, policy, and tooling across justice sector institutions in 30+ jurisdictions.

1,415
Total records
5
Research databases
30+
Jurisdictions covered
March 2026
Last synced

Every governance requirement, procurement clause, and product feature in Brandeis TRP traces to a documented failure pattern or gap identified in these databases. The research is the product specification.

Five databases. One product.

Each database tracks a distinct dimension of AI in justice. Together they map the full landscape that Brandeis TRP is designed to govern.

AI Procurement by Courts

52+ records

Global court AI procurement records

Speech-to-text and transcript workflows dominate heavily across very different jurisdictions and court types.

Insight #60 & #141

AI Controversies in Justice

253+ records

Legal challenges, court cases, adverse findings

"Is AI responsible for part of a decision that affects rights, and can the public meaningfully challenge it?"

Insight #102 - the strategic question driving transparency litigation

AI Adoption Tracker

400+ records

Operational AI deployments across justice institutions

The governance burden grows sharply as AI moves closer to contested decisions.

Insight #41 - the core risk pattern across all institution types

AI Policy Frameworks

300+ records

Regulatory instruments, guidelines, court rules

Courts are carving out a distinct "AI in adjudication" lane: adoption is allowed, but only with human control, transparency, and professional-responsibility guardrails.

Insight #61

NPAI Tool Tracker

400+ records

AI tools deployed across legal and justice institutions

AI adoption in the access-to-justice sector has not been AI-led. AI tools are typically adopted as part of a wider digitisation process, embedded within other technology platforms.

Insight #46
Methodology

How the research is conducted.

Records are sourced through systematic review of publicly available procurement documents, court filings, regulatory publications, academic research, and journalism. Each record is classified, tagged by jurisdiction and institution type, and cross-referenced against the Stimson Centre governance framework.

The databases are maintained and updated by Nicolas Patrick and published at nicolaspatrick.me/research. They represent the most comprehensive open dataset on AI adoption in the justice sector currently available.

The governance deficit statistics cited on this site - 78.8% zero governance, 253 documented controversies, 9.4:1 controversy-to-procurement ratio - are derived from cross-referencing the procurement and controversies databases against each other and against the Stimson Centre's 10-feature governance framework.

Governance Framework Reference

Stimson Centre Framework

The primary governance reference used to assess procurement records is the Stimson Centre's January 2026 report: AI in Global Majority Judicial Systems. The report identifies 10 governance features necessary for responsible AI deployment in courts.

1Human oversight
2Transparency / explainability
3Bias detection
4Privacy protection
5Verification / accuracy
6AI usage guidelines
7Judicial independence
8Public trust
9External audit
10Scope limitation

Key finding: 78.8% of court AI procurements score zero against this framework. Bias detection and external audit score zero across all 52 records analysed.

The research identifies the problem.
Brandeis TRP is the solution.

Every governance gap documented in these databases is addressed by a specific product feature or procurement clause in Brandeis TRP. The research is not background material - it is the product specification.