Polina Zvyagina

Polina Zvyagina

Executive Lead Counsel General Motors
Polina Zvyagina

Polina Zvyagina is Director of AI & Data Governance and Executive Counsel at General Motors, where she leads enterprise‑wide AI and data governance strategy for products that move through the real world, from advanced manufacturing, embodied AI in vehicles and enterprise level AI adoption. A global leader in AI law, policy, and governance, she brings more than 15 years of experience from GM, Meta, Airbnb, Uber, and Apple, building programs that turn AI safety frameworks into operational reality at scale. Before joining GM, Polina led AI Policy & Governance at Meta, where she helped shape the company's approach to AI trust, open‑source releases like Llama, and global regulatory strategy across the White House, NIST, and international AI safety initiatives. Her work has informed emerging global frameworks for AI transparency, accountability, and foundation model governance, and she has co‑authored widely cited pieces on AI system cards, system‑level transparency, and foundation model governance. Known for building 0→1 governance functions that engineers actually want to use, Polina spends her time at the intersection of law, safety engineering, and real‑world deployment, asking fun questions like: what does "responsible AI" mean when the model is not just answering prompts, but steering two tons of metal down the highway?

Responsible AI Summit North America: Day 2 - Wednesday, June 24, 2026

9:40 AM Morning Presentation – From Point Advice to End‑to‑End Systems: How Product Counseling Shapes Responsible AI

Most responsible AI conversations live at the level of principles and frameworks. In reality, the hardest work happens in the messy middle—inside product reviews, design docs, contract negotiations, and "can we ship this?" meetings.

In this keynote, Polina Zvyagina, Director & Lead Counsel for AI and Data Governance at General Motors, will share how day‑to‑day product counseling across the AI lifecycle—idea intake, model selection, system design, testing, launch, and operations—is being used to design end‑to‑end governance solutions at an enterprise scale.

She will walk through concrete patterns from counseling work in safety‑relevant and customer‑facing AI initiatives, and show how those patterns feed back into AI intake workflows, risk taxonomies, documentation, and technical controls. The session is focused on what general‑counsel and product teams can do today to turn scattered AI advice into a coherent, durable responsible AI system.

Check out the incredible speaker line-up to see who will be joining Polina.

Download The Latest Agenda