LONDON — The rapid adoption of artificial intelligence by major banks could leave the financial system increasingly dependent on a small number of technology companies, creating risks ranging from widespread outages to higher technology costs, a new analysis by Moody’s has found.
Moody’s said AI ultimately could reduce costs and boost revenue for banks and other financial companies, but achieving those gains will require substantial investment. Competition among financial institutions could also erode some of the benefits as firms pursue similar technologies and efficiencies.
Among the risks identified by Moody’s are data privacy, cybersecurity, fraud, rapid deposit withdrawals and growing dependence on a relatively small number of AI model and cloud-computing providers.
‘Systemic Dependency’
“The reliance of most financial firms on a relatively small set of foundation AI model and cloud computing providers risks creating a systemic dependency,” Moody’s said, according to The Guardian.
A disruption at one major technology provider could spread rapidly across financial institutions and other industries, Moody’s warned. The rating agency said regulators are likely to increase scrutiny of operational resilience and the concentration of third-party AI providers as adoption grows.
The concerns are particularly significant given the financial sector’s rapid embrace of the technology. More than 75% of financial companies in the City of London use AI, according to a U.K. Treasury select committee report cited by The Guardian. Banks and insurers are using the technology for administrative functions as well as operations such as processing insurance claims and evaluating borrowers’ creditworthiness.
Additional Vendor Risks
Moody’s also warned of “vendor dependence risk,” in which dominant AI model and infrastructure providers eventually gain greater power to set prices. The risk could increase as AI companies face pressure from investors to generate profits.
Financial institutions have some defenses against that risk, Moody’s said. Large banks and insurers retain control of valuable proprietary data, have extensive experience negotiating technology contracts and can use open-source AI models or partnerships to reduce dependence on individual vendors.
At the same time, AI could have significant implications for bank employees. Moody’s estimated there is a 20% chance that by 2030 AI systems will be capable of performing work equivalent to that of a “solid mid-level employee.”
Funding at Risk
The technology could also affect banks’ funding by making it easier for consumers to identify and quickly move money to accounts paying higher interest rates, potentially accelerating deposit outflows.
“In this context, depositors’ trust in the institution and the resilience and stability of deposit funding are critical,” Moody’s said.
The Guardian reported that Lloyds Banking Group is among the financial institutions increasing its investment in AI. CEO Charlie Nunn recently outlined a £13 billion strategy that includes using AI and other technology to attract business, improve efficiency and increase shareholder returns. The plan includes £2 billion in cost reductions and is expected to affect staffing.
Moody’s said that despite the risks and substantial upfront costs, AI is expected to provide longer-term financial benefits to banks and other financial institutions as the technology becomes more deeply integrated into their operations.
