Studying AI in practice

AI systems are rarely just technical models. They are developed inside organizations, shaped by supply chains, evaluated through practical routines, and governed through policy and standards. Agathe Balayn studies exactly these points of contact – the places where AI becomes part of real-world practice.

At RC Trust, Agathe Balayn will pursue this work within Jat Singh’s Compliant and Accountable Systems group. As a postdoctoral researcher, she brings together questions that are often treated separately: how AI is built technically, how it is used in organizations, and how it is shaped by policy. She is especially interested in how AI systems are produced, evaluated, and governed across the AI supply chain.

This perspective closely connects with the group’s work on accountability, trustworthy AI, and the governance of emerging technologies. Through qualitative research, Agathe Balayn studies organizations that produce or use AI systems. She examines the potential harms that can arise from these practices, as well as the challenges organizations face when working toward more trustworthy AI.

Before joining RC Trust, Agathe Balayn conducted academic research while being embedded in companies such as ServiceNow and Microsoft Research. These settings also served as research sites for her work, allowing her to study AI development and deployment close to practice.

Her recent research looks at evaluation practices for AI and generative AI systems, as well as policy questions around AI explainability, fairness, and transparency. At RC Trust, she is also beginning to study standardization processes for AI, with a focus on how standards-makers understand and govern AI supply chains.

Beyond her academic work, Agathe Balayn enjoys sketching and uses illustration to make topics such as AI and the Internet more accessible.