In this WLDA Deep Tech Talk, Ilana Blumenfeld shares a behind-the-scenes look at how PwC matured its Responsible AI capabilities—from abstract ethical frameworks to enterprise-ready solutions. Drawing on years of experience leading global R&D in Responsible AI, she outlines what it truly takes to build systems people can trust—while navigating the messy, socio-technical realities of AI in practice.
This session is a must-watch for leaders looking to bridge the gap between AI theory and implementation, and who seek to embed responsible design deeply into business operations.
Key Takeaways
- Responsible AI isn’t just technical—it’s organizational.
Ethics can’t be solved with algorithms alone. Embedding Responsible AI requires coordination across legal, risk, engineering, and business functions. - From principles to processes: fair isn’t a checkbox.
Abstract values like fairness and accountability must translate into operational practices—governance, design reviews, testing frameworks, and oversight structures. - AI systems reflect human decisions.
Since AI is socio-technical, bias and risk often originate from how data is collected, interpreted, or applied—not just how models are trained. - Don’t just ask “is this fair?”—ask “fair to whom?”
Fairness is not a universal standard. Responsible AI means understanding impact from multiple stakeholder perspectives and being transparent about tradeoffs. - You need alignment before deployment.
Getting buy-in from leadership and teams early is critical. Responsible AI can’t succeed if it’s bolted on after the fact—it has to be designed in from the start.
Speaker: Ilana Blumenfeld, Director, PwC Digital Assurance & Transparency; Responsible AI R&D Lead, PwC
Duration: 44 minutes