Inquiries

Human Agency in an Age of Intelligent Systems

What people need to keep their judgment and authority as AI takes on more of the work.

In preparationHuman Capability · Technology · Governance

Human oversight has become the default answer to AI risk, while the routine work through which people used to build judgment is being automated. This inquiry asks what replaces those pathways, what strong oversight actually requires, and who decides when an AI system acts on someone’s behalf.

It frames the Tokyo Salon, in preparation for February 2027, which would take the question beyond governance to the broader work of societal adaptation, drawing on Japanese and wider Asian perspectives.

Status
In preparation
Domains
Human Capability · Technology · Governance
Why more than one
The capability in question belongs to people, the systems displacing it are technical, and the protections being written for it differ by jurisdiction.
Convening
Tokyo Salon · In preparation · February 2027
Contributors
Avisikta Sinha, Ashwin Telang, Ronan Amir
Funding
Supported from the Institute’s general funds. Organizations interested in supporting this inquiry can read about partnerships.

Open questions

What replaces the developmental function of entry-level work when AI absorbs routine tasks?Planned format: Research BriefScoping
What does strong human oversight of AI actually require, and who is being trained to exercise it?Planned format: Searchlight seriesDrafting
What does cognitive sovereignty mean at the institutional level, and how do organizations put it into practice?Planned format: Research BriefScoping
How should institutions support mid-career capability renewal and workforce re-entry?Planned format: Research BriefScoping

Early publications

Who May an AI Tell? Escalation Authority Has Become a Product SettingAvisikta SinhaSearchlight 029
October 16, 2026Companies Blaming AI for Layoffs Will Have to Say So on the RecordAshwin TelangSearchlight 027
October 2, 2026Proof of Work: Mathematicians and AIRonan AmirSearchlight 019
August 7, 2026The Verification Gap: The Oversight Skill Nobody Is TrainingAshwin TelangSearchlight 018
July 31, 2026