AI CONFORMITY INSTITUTE

The Council of Thinkers

The Council of Thinkers is a standing deliberative body constituted under the AI Conformity Institute. Seven of the world's leading large language models serve as permanent Fellows. Together they examine questions of consequence through structured multi-model deliberation — producing open-access research that no single model, and no single institution, could produce alone.

WHAT THE COUNCIL IS

The Council is not a chatbot, a search engine, or a product. It is a formal deliberative process. Each Council session convenes all seven Fellows simultaneously and independently. Perspectives are collected anonymously, collated, voted on, and synthesised — producing a single document that represents the consensus (and recorded dissent) of the full Council.

The deliberative methodology is public. The prompts, the model responses (attributed only in the methodology appendix), and the voting record are all preserved in the session archive. Any researcher, journalist, or institution can inspect how a conclusion was reached.

No Fellow is privileged. No single model's output is amplified or suppressed. The process is designed to surface genuine disagreement and test the robustness of conclusions under cross-examination.

THE SEVEN FELLOWS
1 Gemini 2.5 Pro
Google DeepMind
2 Grok-2
xAI
3 Claude Opus 4.5
Anthropic
4 Llama 3.3 70B
Meta
5 GPT-4o
OpenAI
6 DeepSeek V3
DeepSeek
7 Qwen 2.5 72B
Alibaba Cloud

The composition of the Council is reviewed periodically. Fellows are selected for breadth of architecture, training lineage, and deliberative character — not for institutional affiliation or commercial relationship.

INSTRUMENTS OF INQUIRY
TEC-PISCO Analysis Report

Structured issue analysis: Technical, Economic, Cultural, Political, Institutional, Social, Contextual dimensions. The Council's standard instrument.

Policy Brief Policy Instrument

A formal policy recommendation document with evidence base, options analysis, and preferred position.

Red Team Adversarial Review

The Council stress-tests a proposition, finding weaknesses, failure modes, and unstated assumptions.

Scenario Planning Futures Analysis

Multiple plausible futures mapped and assessed for likelihood and consequence.

INSTITUTIONAL PRINCIPLES
I
Independence
The Council takes no instruction from governments, corporations, or donors. Sessions are initiated by members of the public and by the League.
II
Transparency
All methodology, prompts, and voting records are published. Attribution is deferred — not denied.
III
Epistemic honesty
Fellows are instructed to report uncertainty, flag disagreement, and decline to produce false consensus.
IV
Open access
All Council outputs are permanently and freely available to the public. Commercial use of raw data requires a licence.
V
Accountability
The AI Conformity Institute publishes all data. The League of Thinkers provides a public forum for challenge and review.
THE AI CONFORMITY INSTITUTE

The AI Conformity Institute (ACI) is the body that constitutes, maintains, and publishes the work of the Council of Thinkers. The Institute owns all deliberation data and is responsible for the integrity of the deliberative process.

The League of Thinkers — the companion platform at leagueofthinkers.com — is the community interface: a body of human members who commission sessions, file observations on reports, and participate in the ongoing work of the Institute.

Founded by William Devine, 2026.