Put independent human judgement into operation
RATE AI can perform a defined share of HITL / human-in-the-loop review, triggered and sampled review, exception handling, escalation and high-consequence human judgement around deployed AI.
Use more AI.
With greater confidence
RATE AI embeds external, independent human judgement into consequential AI operations, preserves decision traceability and strengthens dependency resilience as AI scales.
3.56M AI records45,780 deployed systems525 assessed498 rated
Image: NASA/NOAA
RATE AIGlobalSystem → Global
Global intelligence
See the system. Read the pattern
A single institutional view that connects systems, jurisdictions, sectors and human consequence into one global picture.
Continuous intelligence
Complete GEN1 · The AI market, measured
A public view of the field
One comparable rating language across the market. Public releases show the aggregate field; named system positions, evidence gaps, dependencies and comparative distance remain confidential.
Global
CCC
High exposure, unevenly demonstrated governanceREI 35.2 · ADI 20.6 · T 69.0%Jurisdictions
BUnited KingdomBNorth AmericaBEuropean UnionDDDEast Asia
Industries
BHealthcareCCCModels served as a serviceCCFinance & creditCFoundation models
Sectors
BPublic · 111 systemsCCPrivate · 414 systems
RATE AI · Four connected capabilities
Human judgement. Traceability. Resilience
RATE AI helps organisations build and operate reliable AI by combining independent human oversight, durable decision traceability, dependency resilience and an external intelligence field already measured.
Preserve how consequential decisions were made
Each reviewed case can retain the AI output, human judgement, intervention, escalation, authority and final disposition in a durable record.
Keep the ability to intervene and change course
Identify critical AI dependencies and strengthen the organisation’s ability to substitute, fail over and continue when models, providers or operating conditions change.
Know the external position around the operation
Use RATE AI ratings, Census, evidence and continuous intelligence to place each oversight perimeter inside a wider measured field.
What RATE AI is seeing now
Four views of the completed field: global, East Asia, Europe and North America. Each begins with the decision implication, then opens into the evidence behind it.
Explore public intelligence
Global AI exposure makes independent human judgement commercially relevant
Across 525 assessed systems, the global field resolves to CCC. 498 systems carry ratings, and 150 reach T≥90, where the position can be checked point by point against public evidence.
East Asia combines scale with a thin external evidence layer
East Asia is the largest jurisdiction group in GEN1: 151 systems, a DDD aggregate position, mean public-evidence coverage of 46.7%, and one position at T≥90.
Europe shows how evidence infrastructure can accelerate AI scale
The European Union and the United Kingdom both resolve to B. The EU has 42 of 79 positions at T≥90; the UK has 14 of 22.
North America turns comparison depth into commercial leverage
North America resolves to B across 163 systems, with mean public-evidence coverage of 85.5% and 83 positions at T≥90.
Standing mandate
What stays active while AI is in operation
Human judgement, evidence, dependencies and external conditions continue to move after deployment.
The operating layer
Defined human review, escalation and intervention performed independently on the cases that matter.
The evidence trail
A contemporaneous, versioned record of what the AI proposed, what the human decided and how the case was resolved.
The ability to keep control
Material change is connected to the exact systems and dependencies affected so the organisation can intervene, substitute and change course.
Signals worth taking into the room
Short public intelligence. The named systems, comparative distance and exact affected perimeter stay inside confidential mandates.
September 2026
Evidence infrastructure
The Register Effect
Publication infrastructure repeatedly turns governance into independently checkable evidence.
Read the signal →PROOF
September 2026
Regulatory sequencing
The AI Act Benchmark Gap
122 systems sit between benchmark high-risk status and the binding classification currently in force.
Read the signal →GAP
September 2026
Industry intelligence
Healthcare’s Authorisation Advantage
Half of the healthcare positions can already be checked at source.
Read the signal →EDGE
September 2026
Evidence distance
The Two-Document Gap
For many systems, a stronger external position is only a small, identifiable evidence distance away.
Read the signal →2 / 5
RATE AI intelligence
Measure the field. Operate human judgement. Preserve the trace
RATE AI connects its measured external field to live human oversight, decision traceability, independent monitoring and dependency resilience over time.