Marsh Says Fed Should Ease Up on Banks That Insure AI Loans By Esteban Duarte and Katanga Johnson | November 12, 2026
Key Takeaways & Highlights
- is urging US regulators to allow banks to hold less capital against loans covered by credit risk insurance as they look for new ways to offload data center exposure.
- “We’ve been doing a lot of advocacy around that in the US,” Larsen said in an interview with Bloomberg News.
- “Hopefully we’ll have a positive answer from the US regulators at the end of the year,” she added.
Key Background and Context
Email This
Subscribe to Newsletter
Latest Developments and Impact Analysis
Article
0 Comments Marsh & Mc Lennan Cos. is urging US regulators to allow banks to hold less capital against loans covered by credit risk insurance as they look for new ways to offload data center exposure. Demand for hedging AI-related finance has exploded, but US lenders have traditionally leaned more heavily on other instruments because insurance doesn’t free up regulatory capital for them, according Azi Larsen, a managing director in the credit specialties practice at the world’s biggest insurance broker.
“We’ve been doing a lot of advocacy around that in the US,” Larsen said in an interview with Bloomberg News. “Hopefully we’ll have a positive answer from the US regulators at the end of the year,” she added. The Federal Reserve is set to complete its latest regulatory capital rules in the coming weeks, Vice Chair for Supervision Michelle Bowman said at an event on Nov 10.
The package is expected to include a plan that would lower the capital surcharge for the biggest US lenders and another that would change risk-based capital requirements for midsize and large banks. Credit risk insurance has long been used by European banks to mitigate losses and release capital that can then be used to issue more loans.
Lenders pay premiums to cover a percentage of losses on specific facilities. Barclays Plc has been looking to buy cover on more than $10 billion of data center debt, while HSBC Holdings Plc has explored a similar transaction. With global spending on data centers projected by Mc Kinsey & Co to hit $6.7 trillion by 2030, debt financing is expected to play a much bigger role, and finding different tools to manage risks has become “mission critical” for lenders, Larsen said.
Marsh is active in arranging credit risk insurance for project finance, including deals related to data centers. Despite its limitations, US lenders are increasingly turning to the product to cap their exposure to certain borrowers or industries as the volume of AI-related lending balloons.
“Even if they don’t get capital relief, they still see value in trying to use it,” Larsen added. Topics Insur Tech Data Driven Artificial Intelligence Was this article valuable? Yes No
Here are more articles you may enjoy. Oil Falls as Trump Signals Restraint on Iran Despite Fresh Shipping Attacks New Florida Property-Tax Cut May Bring an Unwanted Side Effect Allstate’s AI Ecosystem to Include a Personal Agentic Assistant for Customers Five Weeks After Isaias, 68,000 Claims and $736M in Losses, Florida Carrier Data Show
Written By Esteban Duarte
More From Author
Written By Katanga Johnson
More From Author
Interested in Ai? Get automatic alerts for this topic.
Email This
Subscribe to Newsletter
Categories: National News Topics: credit risks, data center risk, data centers Have a hot lead? Email us at newsdesk@insurancejournal.com

Comments 0
Leave a Comment
No comments yet. Be the first to share your thoughts!