AI Verification Initiative
Wider priorities and questions advancing independent AI verification.
JUST VERIFY THINK TANK
Convening the people advancing practical AI verification.
The Just Verify Think Tank is a standing platform for initiative-led, bespoke and collaborative roundtables with organisations, industries, jurisdictions and events, examining AI trust, verification, ethics, governance, agentic AI, organisational control, responsible adoption and responsible reliance.
SCOPEIndependent inquiry
FORMATSRoundtables · Research · Programmes
PARTICIPATIONCurated for the question
DIRECTIONFrom evidence to action
WHY THE THINK TANK EXISTS
No single discipline holds every perspective required to verify AI well.
Its agenda can be informed by the AI Verification Initiative, emerging sector and jurisdictional needs, independent Just Verify inquiry and carefully governed collaboration.
Organisations, sectors and jurisdictions face different AI risks, dependencies, evidence requirements and operating conditions. The Think Tank creates a structured environment in which those perspectives can meet, challenge assumptions and examine what responsible reliance should require in practice.
It connects technology, governance, research, infrastructure, policy and real-world adoption while preserving Just Verify’s independent judgement.
A STANDING PLATFORM
One Think Tank. Multiple pathways into inquiry.
The Initiative is an important source of questions, but it is not the Think Tank’s only route. Independent, sector-specific, jurisdictional and collaborative programmes can also inform the agenda.
Independent Just Verify Inquiry
Emerging questions that require focused examination or a new programme.
Sector & Jurisdiction Needs
Distinct operating, policy, market and infrastructure conditions.
Collaborative & Supported Programmes
Relevant programmes developed with suitable collaborators or sponsors, without transferring editorial control.
CURRENT THINK TANK SERIES
Three cities. Three areas of inquiry. One connected series.
The 2026 AI Verification Think Tank Roundtable Series is one programme within the wider Think Tank. Every city examines Trust, Verification, and Sovereignty & Independence before the perspectives are brought together in one connected synthesis.
CAMBRIDGE
Wednesday
4 November 2026
OXFORD
Thursday
5 November 2026
LONDON
Friday
6 November 2026
EVOLVING AREAS OF INQUIRY
A wider programme for the questions influencing AI trust.
These areas describe current priorities rather than fixed boundaries. Additional themes can be introduced where a programme, sector, jurisdiction or emerging verification need requires them.
- 01
AI Verification
What evidence should organisations require before relying on AI?
- 02
Agentic AI
How should identity, authority, delegation and behaviour be verified?
- 03
Evidence & Accountability
How can claims, decisions and responsibility remain traceable?
- 04
Reliability, Risk & Assurance
How should performance, failure and changing conditions be assessed?
- 05
Sovereignty & Infrastructure
How can organisations preserve meaningful control over external dependencies?
- 06
Responsible Adoption & Reliance
What should organisations know before they adopt, rely on, scale or delegate to AI?
HOW THE THINK TANK WORKS
A structured route from question to continued action.
The format adapts to the issue being examined, but the purpose remains consistent: bring the right judgement to the right question and translate the inquiry into useful direction.
- 01 / FRAME
Define the question
Clarify the verification issue, organisational decision or evidence gap that needs examination.
- 02 / CONVENE
Bring together relevant perspectives
Curate the disciplines, sectors and forms of judgement the question requires.
- 03 / EXAMINE
Test evidence and assumptions
Compare experience, surface tensions and identify what remains unresolved.
- 04 / CONTINUE
Translate inquiry into direction
Develop findings, research priorities, practical work or continued programmes where justified.
WHO CONTRIBUTES
Cross-industry. Cross-disciplinary. Cross-perspective.
Participation is based on what somebody can contribute, not simply their sector or job title. The perspective families below are deliberately broad and are not fixed categories.
Organisations & Enterprise
Leadership, operations, risk, strategy and responsible adoption.
AI, Technology & Product
Systems, agents, data, product, innovation and emerging technology.
Law, Finance & Professional Services
Legal, investment, assurance, advisory and commercial judgement.
THE SHARED TABLEDifferent forms of judgement
Broader participation wherever it strengthens the inquiry.
Government, Policy & Public Sector
Public responsibility, policy, governance and institutional oversight.
Research, Standards & Assurance
Academic, evaluation, certification, testing and evidence perspectives.
Infrastructure, Security & Society
Resilience, privacy, physical and digital infrastructure and public interest.
OUTPUTS & CONTINUED WORK
Discussion is the beginning, not the end.
Think Tank activity identifies what matters, what is supported by evidence and what needs to happen next. Roundtable findings may inform further research, Trust Lab projects, practical pilots, organisational solutions and longer-term partnerships.
- 01
Synthesise responsibly
Capture themes, disagreement, evidence gaps and unresolved questions without manufacturing consensus.
- 02
Set working priorities
Identify where deeper research, testing, methods or organisational work is required.
- 03
Continue the inquiry
Develop further roundtables, research, Trust Lab activity or sector and jurisdiction programmes.
- 04
Translate insight into action
Support practical decisions, verification approaches and collaborative work where appropriate.
PRIVACY, ATTRIBUTION & INDEPENDENCE
Designed for open discussion and independent judgement.
Sponsors and collaborators may support a programme, but they do not control the agenda, participant balance, discussion, findings or conclusions.
Private where required
Sessions may be private or invitation-only to support candid, focused discussion.
Responsible attribution
General themes may be shared, but individual comments are not attributed without permission.
Independent judgement
Participation or support does not create endorsement, certification or control over Just Verify’s conclusions.
WAYS TO ENGAGE
Choose the route that reflects how you want to take part.
Every request is considered against the relevance of the proposed contribution, the needs of current programmes and the independence of the Think Tank.
Bring your perspective
Join a roundtable, inquiry programme or focused working discussion where your experience can strengthen the conversation.
Support a programme
Provide financial support for a series or programme while Just Verify retains control of the agenda, participation and outputs.
Develop relevant work together
Work with Just Verify to develop a bespoke roundtable, connected inquiry or collaborative programme around a strategically important AI trust question.
START A CONVERSATION
Tell us how you would like to engage.
Share a concise overview and the Just Verify team will consider the most appropriate route.
ADVANCE THE NEXT INQUIRY
Bring the question, perspective or support that moves verification forward.
The Think Tank creates space for informed judgement and provides a route from discussion into continued work.