The role of investors in the age of AI - Part 2

In this episode, Cambria Allen-Ratzlaff, Interim CEO of the PRI, is joined by Michael Benedict Yamoah (Vice President, Stewardship Director, EOS at Federated Hermes), Chris Jurgens (Senior Director, Omidyar Network), and Oumou Ly (Non-resident Research Fellow, UC Berkeley Center for Long-Term Cybersecurity) to explore how investors should respond to AI.

Building on Part 1, this episode moves from theory to practice, outlining how investors can assess AI governance, identify risks across portfolios, and begin engaging with companies in a fast-moving and uncertain landscape.

Overview:

AI is already reshaping portfolios, but most investors are still early in understanding how to manage the risks. This episode focuses on practical steps, from governance and engagement to tools, research, frameworks and real-world examples of leading practice.

A key message is that there is no perfect framework yet. Instead, investors must start now, build capability over time, and engage continuously as the technology evolves.

Detailed coverage:

What good AI governance looks likeAt a minimum, companies must comply with regulation and establish clear internal policies. Strong governance goes further, embedding AI into enterprise risk management, assigning board-level responsibility, and ensuring oversight across the organisation.

Beyond compliance: lifecycle thinkingInvestors are encouraged to assess the full lifecycle of AI systems, from development and deployment to real-world impacts, liabilities and societal consequences.

AI risk is dynamicUnlike other technologies, AI systems evolve post-deployment. This requires continuous monitoring, disclosure and adaptation, rather than one-off assessments.

Examples of leading practiceCompanies such as Anthropic and Microsoft are highlighted for transparency, investor engagement and responsible AI frameworks. Across the ecosystem, progress is being driven by collaboration between companies, investors and policymakers.

The importance of infrastructure and ecosystemsAI is not just about software, it spans chips, data centres and energy systems. Managing its risks requires coordination across the full value chain.

Practical starting points for investorsInvestors should map where AI sits in their portfolios, identify key use cases, and assess associated risks such as cybersecurity, compliance and liability.

Tools, frameworks and collaborationA growing ecosystem of resources, from investor coalitions to research frameworks, is emerging to support engagement and analysis.

A marathon, not a sprintAI governance is an ongoing process. Investors must build long-term capability, stay engaged in dialogue, and avoid waiting for perfect solutions before acting.

Start now, signal intentEven simple engagement, asking basic governance questions, can send a strong signal to companies that responsible AI matters.

Chapters:

00:08 - Introduction: from AI risk to investor action01:00 - What good AI governance looks like03:05 - Internal policies, risk management and board oversight05:00 - Lifecycle thinking and real-world impacts08:17 - Examples of leading practice in AI governance10:30 - Defining and understanding AI risk13:15 - Mapping AI use cases across portfolios15:39 - Practical tools and investor resources19:44 - Why AI is a marathon, not a sprint22:24 - Final takeaways: start now and engage

Further reading: Anthropic labor market impacts, Microsoft transpare