The Governance Gap in AI-Driven Portfolios

The OpenAI board crisis of November 2023 was a warning that the governance structures designed for conventional technology companies are inadequate for organisations developing transformative AI systems. Two years on, that warning has been only partially heeded. Across the private equity and family office landscape, AI governance remains an afterthought — a checkbox in due diligence rather than a genuine framework for managing the risks that AI introduces into portfolio companies and investment processes alike.

The governance gap manifests in several ways. At the portfolio company level, boards are increasingly asked to oversee AI systems that their members do not fully understand — systems that make consequential decisions about customers, employees, and counterparties. The liability exposure from AI-driven errors or biases is real and growing, yet most board charters have not been updated to reflect it. At the fund level, GPs are deploying AI in deal sourcing, due diligence, and portfolio monitoring without clear disclosure to LPs about how these tools affect decision-making or where the accountability lies when they fail.

The regulatory environment is tightening. The EU AI Act, which entered into force in 2024, imposes specific obligations on high-risk AI systems — and many applications in financial services fall into that category. Singapore's MAS has issued guidelines on responsible AI use in financial institutions. The direction of travel globally is toward greater transparency, accountability, and documentation of AI decision-making. Investors who are ahead of this curve will find it easier to operate in an increasingly regulated environment; those who are behind it will face remediation costs and reputational risk.

At Lensbridge, we have made AI governance a standard component of our due diligence framework. We ask portfolio companies and fund managers not just what AI tools they use, but how those tools are governed, audited, and explained. It is an area where the quality of management thinking is often more revealing than the financial model — and where the best operators are already building durable competitive advantages.