AI & ML interests

Council of AI (CSOAI Ltd, UK Companies House 16939677). Independent measurement body. We measure AI systems on published, frozen splits, sign the card, and re-attest. We do not certify, accredit, or act as a notified body. HF is a printer of the live board — cite GET https://councilof.ai/api/gspc (totals.public_count = 22 axis · 22 measured; unmeasured = 0). Do not freeze a table here. Lid: 22 axes measured · 14 model fleets · 3 public leader scores · 8 fact runs · TIE is TIE · not a certificate. Verify for free: https://councilof.ai/gspc-verify. Canonical banks: the csoai/gspc-* datasets on the live API / collection “GSPC measurement banks (Council of AI)”. Older *bench names are DEPRECATED. Website: https://councilof.ai

Recent Activity

Nicholastempleman  updated a dataset about 1 hour ago
csoai/evidence-index
Nicholastempleman  updated a dataset about 1 hour ago
csoai/disclosure-completeness
Nicholastempleman  published a dataset about 1 hour ago
csoai/disclosure-completeness
View all activity

Organization Card
Council of AI · CSOAI Ltd

GSPC: the public AI benchmark

Measure the claim. Keep the evidence. GSPC publishes scoped AI and domain measurements with their methods, sources and limits. It is measurement, not certification.

Live source · councilof.ai

Current public board

Retrieved from the live board on 30 September 2026: 23/23 axes measured; 14 model-comparison axes and 9 deterministic-fact axes; 0/14 comparison axes separated a leader (7 TIE, 7 UNTESTED). This is a source snapshot, not a new evaluation. Open the live board for current per-axis details.

Read the source JSON
Researchers and developers

Inspect and reproduce

Explore the axis catalogue, datasets, source records and methods. Preserve each run date and its limits.

Browse the dated axis catalogue
AI and agent builders

Connect through open interfaces

Use the public API, MCP endpoint, OpenAPI description or A2A agent card. Discovery is not permission to execute a tool.

MCP · OpenAPI · A2A card

Teams and governance

Commission scoped work

Agree the system, question, method and evidence deliverable first. Payment buys work, never a favourable score or rank.

Discuss a scoped measurement
How to read GSPC: model-comparison results and deterministic source facts are different evidence types. TIE means a lead was not statistically separated; UNTESTED means no separation result is available. Missing is not zero. CSOAI does not certify, accredit or act as a notified body. The live board is authoritative; this Hub page is a readable entry point. Attribute board data to Council of AI / CSOAI Ltd under CC-BY-4.0; third-party datasets retain their own terms.