In the latest episode of the Responsible AI podcast, we engage in a thought-provoking discussion with Oumou Ly, a senior fellow at UC Berkeley's Center for Long-Term Cybersecurity. Oumou shares her invaluable insights from her time working on President Biden's Executive Order on AI, a landmark initiative aimed at addressing the rapid advancements in artificial intelligence. Her experience highlights the complexities involved in crafting policies that can keep pace with the swift evolution of technology. As Oumou aptly notes, "The pace of regulation just sort of moves much slower than the pace of innovation," underscoring the challenges faced by policymakers in this dynamic field.

The conversation delves into the intricacies of integrating "security by design" into programming, a concept that emphasizes the importance of embedding security measures into the development process from the outset. Oumou discusses the hurdles that developers encounter when trying to balance innovation with security, a task that requires a deep understanding of both technical and socio-technical systems. This approach is crucial in mitigating risks and ensuring that AI technologies are developed responsibly and safely, protecting users and society at large.

Oumou also sheds light on the ongoing battle against ransomware, a pervasive threat in the digital age. She explores the dilemmas faced by organizations when dealing with ransomware attacks, emphasizing the need for proactive measures and robust cybersecurity practices. Her insights provide a comprehensive overview of the challenges and opportunities in AI regulation and cybersecurity, offering listeners a glimpse into the future of these critical areas. Tune in to the Responsible AI podcast for an enlightening discussion that navigates the complexities of AI policy, security, and innovation.

Chapters

02:09 The Executive Order on AI: Process and Development

07:30 Major Accomplishments of the Executive Order

10:02 Cross-Disciplinary Collaboration in AI Policy

13:09 Challenges in Legislative Regulation of AI

16:19 Incentives and Interests in AI Regulation

17:01 Socio-Technical Factors in AI and Cybersecurity Risks

19:47 Security by Design in AI Development

21:18 Navigating AI Risk Management

22:21 Ransomware: Strategies and Metrics

24:47 The Ransomware Dilemma: To Pay or Not to Pay?

26:36 Secure by Design: A Holistic Approach

27:10 Cultivating a Culture of Cybersecurity

30:20 Oumou’s work on the Blue Sky Social Network

33:08 Investor Influence on Cybersecurity Practices

36:37 Accountability in AI: Who's Responsible?

38:37 The Importance of Digital Literacy