Abstract:
As AI systems increasingly shape our personal, professional, and societal lives, the question is not only what machines can do, but who controls the values and outcomes they produce. This talk examines both individual agency — the capacity to think, judge, and act — and collective agency, where communities define norms, resist imposed standards, and guide AI deployment. Drawing on research in trustworthy AI, decolonial alignment, and human–AI collaboration, I will explore technical and governance approaches that preserve human autonomy, including transparency tools, scoped alignment methods, and collaborative task structures. I will introduce AI platform cooperatives as a counterweight to tech‑company dominance, fostering community ownership, shared governance, and technological self-determination. Ultimately, AI should be a tool that empowers humans, singly and together.
Kush R. Varshney is an IBM Fellow based at the T. J. Watson Research Center where he is responsible for innovations in AI governance and leads research on trustworthy and purposeful AI. He and his team have created open-source tools such as AI Fairness 360, AI Steerability 360, and Granite Guardian. He independently published the book Trustworthy Machine Learning in 2022.
