Private Credit, AI, and the Future of Insurance Risk

Listen and follow the ‘Risky Science Podcast’
Apple Podcasts | Spotify | Overcast

Two powerful forces are transforming how insurers think about risk: the accelerated growth of private credit and the rise of advanced analytics and generative AI.

In the latest episode of the Risky Science Podcast, we speak with Carmi Margalit, Life Insurance Sector Lead at S&P Global Ratings, about how these dynamics are reshaping investment strategies, risk modeling, and long-term credit profiles across the industry.

🧮 Private Credit: A Rising Share with Risk


Private credit has become a central theme in insurer portfolios, particularly in the hunt for higher yields amid stable but elevated interest rates. According to Margalit, approximately 46% of the industry’s collective bond portfolio is now in private placements, mostly investment-grade 144A securities. But the more complex shift is happening beneath the surface, in what Margalit calls “indirect exposure to non-investment grade private credit,” including asset-backed finance and fund structures.

“It’s not good or bad in general,” Margalit explains. “There’s yield pickup and better covenants, but it comes with complexity, illiquidity, and a perceived lack of transparency from the outside.”

While this exposure remains relatively small—estimated at under 10% of total bond holdings—it raises important questions about liquidity planning and capital adequacy. S&P’s methodology accounts for this through detailed stress-testing and proprietary capital models, and Margalit emphasizes that liquidity is a critical modifier in their rating framework.

🤖 Risk Modeling Meets Machine Learning

Beyond the balance sheet, life insurers are also exploring how AI and advanced analytics can enhance risk forecasting—particularly in mortality and morbidity modeling. Margalit draws a clear line between the operational applications of AI (e.g., streamlining call centers, processing legacy policies) and its use in actuarial science.

While some insurers are cautiously experimenting with large language models and machine learning to refine their mortality projections, Margalit cautions that the industry is still in the early innings.

“There’s growing consensus that machines can be better at pattern recognition,” he notes, “but firms are keeping a human in the loop to avoid the black-box problem—and regulators are starting to ask questions too.”

🔍 Listen to the Full Episode

📄 Show Notes & Resources

1. North American Life Insurers Sector View 2025: Very Cautious Optimism

  • S&P’s stable sector outlook, interest rate environment, private credit growth, and offshore reinsurance use.

2. Private Credit Casts a Wider Net To Encompass Asset‑Based Finance and Infrastructure

  • Explores how private credit is expanding beyond middle-market lending, including into ABS and infrastructure.

3. S&P Global Ratings – Life Insurance Sector Hub

  • Centralized resource for research on the life insurance industry, including credit conditions, sector outlooks, and capital strength.

4. S&P Global – Generative AI Topic Hub

  • Access webinars, newsletters, and AI-specific insights across sectors, including insurance and financial services.