Flagship research programme
As machines gain agency inside organisations, how do we ensure they act as intended?
Who Assures the Agents? convenes enterprise, audit, governance, technology, behavioural and regulatory perspectives around a defining operational question.
The research question
Reliability cannot mean successful task completion alone. Agents require defined roles, boundaries, context, permissions, checkpoints, escalation paths and evidence.
An open programme
The programme is being shaped through research papers, executive interviews, conversations, roundtables and newsletter editions. The programme will publish its methodology alongside future findings.
Participate
The Institute invites leaders and researchers working on enterprise AI, agent governance, operational risk, human oversight and behavioural systems to contribute to the inquiry.
Begin with one workflow
Contribute to the question
Share research, practical experience or a perspective that can strengthen the field.