bank of america The bank’s CEO, Brian Moynihan, told CNBC on Wednesday that it has hired a large number of outside contractors to help patch software vulnerabilities exposed by advanced artificial intelligence models, including Anthropic’s Mythos.
In an interview after the bank’s third-quarter earnings release, Moynihan said the company ramped up broad enhancements to its internal and vendor code earlier this year, leveraging thousands of employees and external talent.
“That risk is a real risk, and that’s why when the Frontier-type model came out. … Many of us sat down and said, ‘How can we find a way to get through this quickly?'” Moynihan told CNBC’s Leslie Picker.
“We hired 1,000 contractors in the second quarter to help us with our patching efforts, in addition to the 3,000 people we have in the region who are still here helping us,” he said.
Moynihan told analysts earlier Wednesday that the bank spent “over hundreds of millions of dollars” in the third quarter on regulatory issues related to software remediation and anti-money laundering compliance.
The comments are the first time a major U.S. company has disclosed the manpower and money spent protecting its software systems from Mythos, which was released to a small number of banks, technology companies and government agencies in April.
It also highlights the dual nature of AI. While AI can help improve productivity and save companies money, the ability of models like Mythos to discover and exploit flaws in decades-old software can also force companies into a costly scramble to defend themselves.
“We have to stay ahead of them and remember that the people who are doing the software development are the same people who are trying to sell us services,” Moynihan said.
Still, Moynihan, speaking as the company’s shares fell as much as 8% on Wednesday, sought to reassure investors that the increase in spending was temporary.
“Once you patch it, you don’t have to patch it as much. … The pipeline will pass,” Moynihan said, adding that the bank expects remediation costs to decrease over time.
Elsewhere, AI is helping to reduce costs. Moynihan told analysts that Bank of America has about 130 internal AI projects and that savings from those efforts are expected to grow from $300 million this year to $700 million by 2027.
He said Bank of America has cut about 5% of its management positions, resulting in a decline in its headcount of about 3,400 people over the past year. It is unclear how much of the workforce reduction will be due to AI.
