‘AI is our friend’: Bank CEOs play down risks from Mythos

Processing Content

• Key insight: During earnings calls this week, bank CEOs addressed AI-related risks and rewards. Most continue to lean heavily into the technology.
• What’s at stake: At the same time, bankers say they are aware of the inherent risks in AI.
• Forward look: Citi plans to outline its AI strategy at an investor day on May 7.

Bankers are continuing to embrace artificial intelligence as an imperative instrument in their toolboxes, even as they acknowledge that new AI models could heighten cybersecurity risks.

On Wednesday, Morgan Stanley CEO Ted Pick laid out his case for AI, saying it’s “just the latest generation of technology that is going to be part of the ecosystem” of the banking industry. In fact, it’s beginning to evolve from a pure efficiency tool that automates routine tasks to a “productivity phenomenon,” Pick told analysts during the bank’s first quarter earnings call.

The $1.6 trillion-asset investment bank is one of the companies that’s currently using a beta version of Anthropic’s Claude Mythos AI model, Pick said. Last week, Anthropic said it was withholding public release of the AI model that detects security vulnerabilities in software, warning that bad actors could exploit coding vulnerabilities faster than banks can fix them.

“AI is our friend … and we’re at an important moment,” Pick said on the call. “This is something we consider to be additive to what we have, which is world-class technology, world-class cyber defense and then the best-trusted advisers sitting with the client. That is … the secret sauce.”

AI usage has become one of the most dominant themes this year, in banking and elsewhere. Supporters point to productivity improvements, problem-solving capabilities and efficiency. Opponents have warned of job losses, data-privacy concerns, security mishaps and ethical dilemmas.

American Banker’s 2026 AI Talent Shift Survey found that at least half of the 206 bankers surveyed in March include AI usage as a high organizational priority. Of the respondents, a broad majority of institutions had increased their AI tech spending by at least 10% over the last 12 months.

Last week, Pick and several other big-banks CEOs, who were already in Washington, D.C., for a Financial Services Forum meeting, went to the White House to talk to Anthropic executives, Federal Reserve Chair Jerome Powell and Treasury Secretary Scott Bessent about potential cyber-attack risks involving Claude Mythos Preview.

Anthropic has restricted access to the Mythos Preview model, sharing it with a consortium of more than 40 technology and critical infrastructure companies, the AI company said Tuesday. A Morgan Stanley spokesperson said Wednesday that the investment banking giant is not part of the group.

AI has come up during all of the big banks’ earnings calls this week. Citi CEO Jane Fraser said her bank is looking at AI through four lenses, including long-term workforce implications, revenue generation, productivity improvement and defensive capabilities such as protection against cyber fraud.

The bank plans to offer more details about its AI strategy at its investor day in May.

JPMorganChase is also testing Claude Mythos, CEO Jamie Dimon told analysts Tuesday.

“While we’re trying to get the benefits of AI, we also are very cognizant of the risks of cyber,” Dimon said on JPMorgan’s earnings call. “I’d say the banks in total are rather well-protected.”

At Morgan Stanley, bankers are prepared to respond to AI risks, Pick said.

“If the ecosystem risk is likely increasing because of the quality and muscularity of the model, then we do need to get our gloves up and take it to another level, and that’s exactly what you’d expect, and we very much intend to do so,” Pick said.

“But I want to say on the back end, that a lot of the good that AI is going to bring both as an efficiency and effectiveness matter should not get dismissed, because that’s an important phenomenon that’s going to continue to transform this firm.”

Also on Wednesday, Bank of America CEO Brian Moynihan pointed to AI as a potential tool to further cut the company’s headcount, which fell slightly year over year to 212,134 people at the end of March.

During a call to discuss the bank’s first-quarter results, one analyst asked how Bank of America is reducing its workforce over time. Moynihan said attrition is one strategy, and so too is the adoption of technology.

“It comes from eliminating work and applying technology … and AI gives us places to go we haven’t gone,” Moynihan said, adding that 200,000 Bank of America employees have access to AI or use it every day.