REGISTER | ADD TO CALENDAR | PERMALINK
On Thursday, 20 August 2026 at 13:00 EDT (17:00 UTC), the Data & Society Research Institute hosts a webinar ‘The Limits of Human Oversight in the Age of AI Agents‘, tied to its AI on the Ground program.
AI agents have moved from speculative promise into everyday deployment, translating high-level goals into plans and carrying them out across connected digital environments. Agent builders often promise that users remain in control through approval prompts and pause buttons, but those safeguards work only when users can recognize a mistake and intervene before consequences cascade.
Drawing on fieldwork in a computational biology laboratory, the Data & Society primer ‘The Oversight Fallacy: Why AI Agents Require More than Humans-in-the-Loop‘ examines what effective oversight of AI agents requires in practice. Co-authors Ranjit Singh and Samir Passi, joined by Margaret Mitchell and Madeleine Clare Elish, will discuss why putting a human in the loop is insufficient and what builders and deployers must do to make oversight meaningful. Closed captioning will be provided.
SPEAKERS
Ranjit Singh – Program Director, AI on the Ground, Data & Society
Samir Passi – Social Scientist and Affiliate, Data & Society; contracted AI Researcher at Microsoft via KForce
Margaret Mitchell – Chief Ethics Scientist, Hugging Face
Madeleine Clare Elish – Head of Evaluations, Responsible Development and Innovation, Google DeepMind
#AIAgents #HumanOversight #AIGovernance