Roundtables

Small working conversations of six to twelve people under the Chatham House rule. There are no presentations. Each one works on three things: what can be done today, what is blocking progress, and which collaborations or resources would make the difference.

Four fellows each led a roundtable in October 2026

All four were held online, and each was designed and led by a fellow of the Sentient Futures Project Incubator. Write-ups are in preparation and will be linked here as they are published.

Led by Pooja Manvikar

Who Should Have a Say in AI’s Future?

Who needs to be involved now, and who sets the terms outside the AI labs?

Courts, health systems, election bodies, insurers, unions and local governments run much of the world, yet few have any plan that assumes AI systems more capable than the people overseeing them. The roundtable brought together people who have built participatory processes, and people who could make the case for one where none exists, and asked who should decide how the interests of AI systems are weighed, if they have them.

Write-up in preparation.

Led by Alejandra Leyva

Human Value and Claims in a Post-Control Scenario

If humans are no longer economically necessary and cannot reliably enforce their demands, what could make their claims to rights, resources and influence meaningful?

Eight participants from different academic and professional backgrounds looked for foundations of human value beyond economic productivity, separated recognizing human interests from protecting them, and asked when personal and collective autonomy could stay meaningful.

A research report by Alejandra Leyva, drawing on this roundtable and the one on 20 October, is in preparation.

Led by Aleksandra Czuchaj

AI Abundance Without Dependency

What institutions and international arrangements should we build today to share the benefits of AI globally without locking low- and middle-income countries into new forms of dependency?

Compute access can be conditional, API access can be repriced or withdrawn, and revenue transfers can leave ownership unchanged. The roundtable compared concrete benefit-sharing arrangements and asked which rights, institutions and exit options must come with access, and how those arrangements would change if advanced AI systems became counterparts or claimants themselves.

Write-up in preparation.

Led by Gagana M D

What Makes Cooperation Without Control Work?

What would make cooperation between humans and advanced AI systems possible and trustworthy when humans can no longer reliably control them?

A follow-on to the 15 October roundtable. Eight participants worked through shared values, common goals, trust, negotiation and credible commitments, and how cooperation could stay reliable as capabilities and interests change.

An essay by Gagana M D is in preparation.

In person: So, What Do We Do Now?

On 29 October, 19 people spent an afternoon in San Francisco on one question: what should be in place before AI systems become too capable to simply direct? They left with four projects and 39 written offers of help.

What came out of 29 October

Each conversation should leave something others can build on

The roundtables are meant to seed work that continues after each session, and to turn separate conversations into a shared agenda others can examine and challenge. We plan to publish:

  • Synthesis notes and policy briefs from each session
  • A working map of the emerging post-control field
  • Connections and follow-up projects that grow out of the series

Take part in a roundtable