Moody's says banks racing to bolt AI onto their operations have made themselves dependent on a small cluster of tech firms, and that dependence is now a credit risk in its own right.
The rating agency warned this week that banks and insurers are leaning on the same few providers, OpenAI, Microsoft, Google and Amazon, for the foundation models and cloud systems that increasingly sit under their business. That's a lot of risk in very few hands. An outage at any one of them, Moody's said, could spread across customers and sectors within hours.
According to a Guardian report on Moody's warning, the agency described the problem as vendor dependence risk. It isn't theoretical. A UK Treasury committee report published on January 20 said more than 75% of UK financial services firms were already using AI, with the heaviest take-up among insurers and international banks. If you're a bank director, that figure should change the way you read the whole AI story. This is no longer a lab tool or a back-office experiment. It's becoming part of the operating floor.
Here's the part that should worry a bank's risk committee more than its IT department. Moody's flagged that AI adoption is adding to older risks, not replacing them: data privacy failures, cybersecurity gaps, fraud and rapid deposit flight. None of that is new. Only the speed is. AI tools make it easier for a customer to compare rates and move money in seconds. If a bank's app, or a rival's chatbot, pushes depositors toward a slightly better yield elsewhere, large pools of cash can move before the old warning lights even come on.
Trust moves fast now.
Then there's price. The Guardian reported that Moody's pointed to the bargaining power of dominant AI providers as they chase profitability after years of heavy losses. A bank that has rebuilt underwriting, fraud detection or customer-service workflows around one foundation model doesn't have many easy exits. Switching providers isn't like changing a stationery supplier. It means re-testing, re-certifying and often re-training systems that regulators expect to be explainable.
The same suppliers carry their own strain #
This isn't Moody's first AI warning this summer. In late July, the agency warned that unprecedented capital spending on AI infrastructure was testing the credit quality of Microsoft, Amazon, Alphabet, Meta, Oracle and CoreWeave. Forbes, citing Moody's analysis, reported that rising capital intensity, debt levels and off-balance-sheet commitments were becoming a credit issue for at least some of the largest technology companies. Other reports put the group's direct debt at about $460 billion, alongside lease commitments well above $1 trillion.
Put those warnings side by side and the picture is not comfortable. The same companies building the AI infrastructure banks want to use are also spending heavily to keep that infrastructure ahead of demand. Banks are not just buying software from suppliers. They are tying themselves to companies whose own balance sheets are being reshaped by the AI race.
Moody's isn't telling banks to stop. The agency still expects AI to cut costs and lift revenue across Wall Street and the City, and that part is easy to understand. Call centers, compliance reviews, fraud checks, loan files, you name it, banks can see the savings. The warning is about order. Adoption is moving faster than the resilience planning meant to sit underneath it.
Banks are still pushing faster #
You can see that speed in the wider financial technology push. The Wall Street Journal reported this week that Wells Fargo plans to roll out tokenized deposits for corporate and commercial clients this fall, representing dollars and pounds as digital tokens on its own blockchain platform. JPMorgan says its Kinexys platform has processed more than $3 trillion since inception and averages more than $5 billion a day. The Clearing House has also said 17 major financial institutions will join a tokenized deposit settlement network targeting the first half of 2027.
None of that is the same technology Moody's is warning about. But it shows the same instinct. Banks are rebuilding the plumbing around speed and always-on automation, while the question of single points of failure becomes harder to dodge.
Frankly, that's the tension Moody's has put in front of the market. Banks spent years building compliance teams to police their own risk. Now they're outsourcing a growing share of decision-making and infrastructure to vendors they don't control and can't easily replace. Regulators are already watching third-party concentration in cloud services under UK and EU rules. AI vendors will not sit outside that perimeter forever.
Also read: Singapore at 61: AI Could Make SG Less Needed for Global Companies, and Challenge the Economic Model That Made It Rich • AI Agents Are Letting Startups Cut Customer Support Headcount in 2026 • Eliza Labs Founder Declares ElizaOS Token Completely Dead After Settlement