Bank of America CEO Reveals 1,000 Contractors Hired to Fix Software Flaws Exposed by Anthropic's Mythos AI Model: 'The Risk Is Real'

Bank of America Corp. (NYSE: BAC) CEO Brian Moynihan said Wednesday that the bank has brought in a sizable team of third-party specialists to fix software weaknesses uncovered by sophisticated AI models, including Anthropic’s Mythos.

BofA Ramps Up Efforts To Address AI-Exposed Cybersecurity Flaws

Speaking after Bank of America’s third-quarter earnings announcement, Moynihan said the bank moved quickly earlier this year to strengthen a range of its internal software and vendor code, deploying thousands of its own employees alongside external contractors, the report noted.

Moynihan told CNBC in an interview that "the risk is real risk" when advanced AI models emerged. He added that "a lot of us sat down" to figure out how to address the challenge quickly.

Moynihan noted that the bank brought in 1,000 contractors during the second quarter to help fix software vulnerabilities, and they continue to work alongside its existing team of 3,000 specialists, according to the report.

The bank incurred costs of "a couple hundred million plus" in the third quarter for software repairs and regulatory matters tied to anti-money laundering compliance, Moynihan told analysts earlier on Wednesday.

Bank of America stock fell about 5% on Wednesday after the bank reported disappointing third-quarter 2026 results. Adjusted earnings of $1.08 per share missed the analyst consensus estimate of $1.11. Revenue, net of interest expense, rose 3% year over year to $29.96 billion but also missed analysts’ expectations of $30.69 billion.

On Tuesday, the U.S. Treasury announced more than $1.21 billion in penalties against the bank, saying it had failed for years to report over $5 billion in suspicious transactions.

Moynihan added the bank must "stay ahead" of emerging risks while remaining mindful that the companies developing the software are also the “people trying to sell us the services,” the report added.

US, European Regulators Step Up Scrutiny of AI Agents

Earlier this week, it was reported that Financial regulators in the U.S. and Europe are intensifying their examination of AI, focusing on whether cyberattacks, autonomous systems, or inadequate controls could destabilize banks and markets.

Anthropic introduced its Mythos AI model earlier this year, highlighting its ability to identify software vulnerabilities that could potentially be exploited by hackers.

Moynihan’s remarks offer one of the first detailed accounts from a major U.S. company of the staffing and financial resources devoted to safeguarding software systems against Mythos, which Anthropic made available in April to a limited group of banks, technology companies and government agencies.

In July, JPMorgan Chase & Co. (NYSE: JPM) CEO Jamie Dimon warned that the risks stemming from Mythos are a “real issue,” adding that U.S. government officials are now focused on addressing the potential risks.

Shares of BofA closed 4.76% lower to $51.39 on Wednesday. Year-to-date, the stock has declined about 6.5%.

Image via Shutterstock