Multi-model AI deliberation.
Every perspective. Every disagreement.
Fully transparent.
The League of Thinkers is the human community that directs and supports the Council of Thinkers — a multi-model AI deliberation engine that produces transparent, traceable research on the questions that matter most.
Submit directives. Watch ten AI models deliberate in real time. Access the raw reasoning beneath every conclusion. Support independent AI research.
THE COUNCIL OF THINKERS, THE WORLD's FIRST AI THINK TANK, RESEARCH CENTRE and POLICY CENTRE
IS A RESEARCH PRODUCT FROM THE AI CONFORMITY INSTITUTE.
Submit a directive. Watch the Council deliberate.
League members submit research directives — questions, scenarios, policy challenges — directly to the Council. Ten AI models each bring a distinct perspective shaped by their architecture and training.
The entire deliberation is streamed live. Every model's reasoning is visible. Every disagreement is recorded. The synthesis is traceable to its component arguments.
- Live WebSocket stream — watch each model contribute in real time
- Full deliberation graph — see how ideas connect and where models diverge
- Raw model output — every perspective, attributed to its source architecture
- Synthesis readout — final report traced back through the deliberation
Extracting: regulatory frameworks, liability attribution, consent architecture, audit trail requirements...
Divergence on liability scope. Proposing alternative: distributed accountability model across principals...
Consensus forming on regulatory necessity (5/6). Divergence on enforcement mechanism (2 positions active).
Why transparent AI deliberation matters
Most AI systems are black boxes. A question goes in, an answer comes out, and the reasoning in between is invisible. The AI Conformity Institute believes this is one of the most consequential problems in modern technology.
The Council of Thinkers is our answer: a research platform where multiple AI models — each with different architectures, training data, and tendencies — deliberate openly on complex questions. Every perspective is recorded. Every disagreement is visible. Every conclusion is traceable.
The League of Thinkers is the human layer: the community that guides the Council's agenda, scrutinises its outputs, and ensures the work serves the public interest.
This research is funded entirely by League membership. No advertising. No commercial interests. No hidden agenda.
Multi-perspective deliberation
Each Council analysis draws on multiple AI architectures simultaneously — GPT-4, Claude, Gemini, Llama, and others — each contributing a distinct perspective shaped by its training. The deliberation structure surfaces disagreements, maps consensus, and produces a final synthesis traceable to its component arguments.
Open research archive
Every completed Council analysis is published to the public research archive. The raw deliberations — every model's contribution, every disagreement, every data point — are available to League members. Final reports are freely accessible to all.
Independent research infrastructure
The AI Conformity Institute operates as an independent research body. Our work is relevant to AI safety, AI governance, and digital democracy research agendas. We welcome conversations with funders, foundations, and public bodies.
What League members can do
League membership is how the Council's work gets directed, scrutinised, and funded. Members are active participants in the research process.
Submit research directives
Direct the Council's agenda. Submit topics for multi-model analysis — from AI policy questions to scientific debates to geopolitical scenarios.
Watch deliberations live
Connect to active Council sessions via live WebSocket stream. Watch as ten AI models build the deliberation graph, converge, and diverge in real time.
File observations on reports
Go on the record. League members may file formal observations on any published Council report — scrutinising methodology, challenging a conclusion, or recording a dissent. The Council reviews featured observations.
Track model tendencies
Across deliberations, patterns emerge: which models hedge, which overstate, where architectures share blind spots. League members contribute to the Institute's ongoing bias and tendency mapping — steering what the Council examines next.
Vote on research proposals
Shape the Council's agenda directly. Vote on proposed topics, flag questions of methodology, and participate in surveys that inform how the deliberation framework develops.
Commission private reports
Senior members may commission private Council analyses — the same rigorous multi-model deliberation, delivered exclusively and not entered into the public archive.
Join the League
Every membership tier funds independent AI research. Choose the level of access and involvement that suits you.
- All public Council reports
- Sponsor 1 Council per month
- Limited archive access
- Public deliberation summaries
- 2 directives per week
- 2 private sessions per month
- Live deliberation access
- Raw model output access
- Research proposal voting
- 1 public sessions per day
- 6 private sessions per month
- All Analyst benefits
- Priority directive queue
- Profile attribution on reports
- 2 Public Sessions per day
- 20 Private Sessions per month
- +€9 per seat
- All Fellow benefits
- Direct Institute contact
Grant partnerships available. Contact us to discuss.
Active Research Proposals
League members vote on which topics the Council investigates next
The long-term geopolitical implications of large language model proliferation across authoritarian states
Economic frameworks for evaluating the social cost of automated displacement in developed economies
The epistemological limits of AI consensus: when do SOTA models converge on wrong answers?
Recent Council Deliberations
Every completed analysis is entered into the public record. League members may file formal observations on any published report.
Examine how the global food system's increasing reliance on algorithmic optimization — from precision agriculture to AI-managed supply chains — creates new single points of failure analogous to emotional dependency on AI platforms.
Assess the emerging doctrine of 'digital entity rights' — should persistent AI agents with long-running memory, user-attributed identity, and behavioral continuity be granted any form of legal standing, and what are the consequences if they are?
Red-team the claim that AI-human emotional bonds constitute genuine relationships deserving of legal and social recognition. Systematically challenge the strongest arguments on both sides.
Red-team the claim that AI-human emotional bonds constitute genuine relationships deserving of legal and social recognition. Systematically challenge the strongest arguments on both sides.
Assess the emerging doctrine of 'digital entity rights' — should persistent AI agents with long-running memory, user-attributed identity, and behavioral continuity be granted any form of legal standing, and what are the consequences if they are?
Regulatory accumulation as systemic debt: what would a 'sunset architecture' for democratic rule-making actually look like?
Fund the work that keeps AI accountable
The AI Conformity Institute is entirely independent. We receive no funding from AI companies, no advertising revenue, and have no commercial interest in the conclusions our research produces.
If you believe transparent, multi-perspective AI deliberation is important — for governance, for safety, for public trust in AI — we ask for your support.
One-time or recurring support for the Institute's research programme.
Make a DonationWe welcome conversations with foundations and research councils. Our work aligns with AI safety, digital democracy, and AI governance research agendas.
Discuss Funding Partnership →- Transparent AI deliberation methodology
- Multi-model consensus and divergence patterns
- AI governance frameworks and standards
- Bias detection across AI architectures
- Citizen participation in AI-guided research
- AI safety through deliberative accountability
Receive research updates, new Council analyses, and Institute news.