Welcome to the Age of “Principled AI”

Everything is coming up AI these days, including me. Last week, I had the privilege of moderating “The Age of Principled AI,” a panel hosted by Fast Company and Booz Allen Hamilton exploring the ramifications of President Biden’s Executive Order and how to regulate AI going forward.

I was joined onstage by Navrina Singh, CEO of Credo AI; John Beezer, senior advisor to the U.S. Senate Committee on Commerce, Science, and Transportation; the Department of Defense’s Dr. Matt Johnson; and Randal Meyer, chief counsel and legislative director to Representative Nancy Mace. According to GPT-4’s summary of the event, here’s what we talked about:

1. Turning Point for AI Ubiquity: Susan Penfield, Booz Allen’s CTO, opened the event, highlighting the White House’s executive order on AI and the growing ubiquity of AI in various sectors. She raised critical questions about AI governance and mitigating unintended consequences.

2. Government’s Role in AI: The panelists agreed on the need for a robust regulatory framework for AI. The discussion centered around the recent executive order, highlighting the need for a coherent national policy and its impact on various sectors, including national security.

3. AI in Legislation and Policy: Discussions involved congressional perspectives on AI regulation, touching on the need for statutory authority for AI governance. The panelists debated approaches to AI legislation, including self-certification, third-party audits, and the necessity of establishing clear standards for audits and assessments.

4. Operationalizing AI Ethics: Matthew Johnson, Senior Technical Advisor in Responsible AI at the DoD, discussed the DoD’s approach to operationalizing AI ethics, including a toolkit for responsible AI deployment. This toolkit aims to provide a framework for aligning AI applications with ethical principles and values.

5. Impact of AI on National Security: The panel addressed the implications of AI in national security, including the importance of using AI responsibly and ethically in defense contexts. They emphasized the need for AI technologies that align with democratic values and principles.

6. Role of Procurement in AI Deployment: Discussions highlighted the role of procurement as a tool for influencing the development and deployment of AI technologies. The need for clear criteria and benchmarks to ensure alignment with ethical principles was emphasized.

7. Innovative AI Use Cases in Government: Examples of AI applications in government agencies, such as the Social Security Administration, were discussed to illustrate the potential of AI in improving services and decision-making processes.

8. Need for AI Literacy and Expertise: The panelists underscored the importance of AI literacy and the need to build capacity within government agencies and other stakeholders to understand and effectively use AI.

9. Challenges in AI Regulation: The challenges of regulating AI were a central theme, with discussions around balancing innovation with regulation, anticipating threats, and ensuring that AI applications do not infringe on rights or freedoms.

10. Future of AI Governance: The event concluded with reflections on the necessity of a multi-stakeholder approach to AI governance, the importance of aligning AI with societal values, and the potential of AI to transform government missions and societal outcomes.