In this episode, Workday's Chief Responsible AI Officer Kelly Trindel sits down with Kasia Chmielinski, founder of the Data Nutrition Project, to get real about what it actually takes to build and use AI without losing human dignity in the process. They dig into: Why "cognitive surrender” is the biggest risk leaders face right now, not the technology itself How anthropomorphizing AI can help or quietly erode trust and dignity, depending on how it's designed What "human in the loop" actually means once agentic systems are making decisions at scale Why data quality, not data volume, determines whether AI works—and why documentation is becoming non-negotiable How regulation like the EU AI Act is impacting real-world AI governance Why organizations need both org-level principles and system-level guardrails to deploy AI responsibly If you're a leader trying to figure out how to keep critical thinking alive on your team while adopting AI at scale, this conversation is for you. 🎧 Subscribe to the Future of Work podcast for more conversations with the leaders shaping how we work with AI. 0:42 — Kasha’s background and path to responsible AI 2:08 — Kelly’s background in psychology and governance 3:32 — Defining responsible AI and agentic AI 5:12 — Responsible AI as a socio-technical approach 7:07 — What’s different about AI right now 9:50 — Human dignity and the risk of cognitive surrender 11:56 — Augmenting expertise instead of replacing it 13:17 — Anthropomorphizing AI 16:12 — Using AI to prepare, but not to flatter 17:25 — Design decisions and public responsibility 18:42 — AI governance in practice 19:41 — Regulations and standards 21:01 — Governance at the organizational level 22:57 — The power of defaults and accountability 23:23 — Managing teams of people and AI 24:20 — Final takeaway: reject techno-determinism Please subscribe and follow us: YouTube: https://www.youtube.com/user/workdayvideo Podbean: https://workdaypodcasts.podbean.com/ LinkedIn: https://www.linkedin.com/company/workday/ Instagram: https://www.instagram.com/workday/

 82