Bank of America's top boss just added his voice to a growing chorus of concern about a powerful AI tool.
Earlier this year, Anthropic introduced Mythos, claiming the model is particularly skilled at identifying network security flaws.
"It is a big change in the amount of work that will have to go on and the pace of which these tools will affect vulnerabilities in your systems, and how fast you have to take care of that," Moynihan said.
He made it clear that the industry is not sitting still. "We are all working hard on it," he added.
Get the market news that matters in a five-minute read with Market Briefs, our free daily newsletter
The financial industry is particularly vulnerable because its systems handle vast amounts of sensitive data and high-value transactions. A breach facilitated by an AI vulnerability scanner could lead to catastrophic losses, which is why executives are prioritizing defensive measures. This has prompted banks to invest heavily in cybersecurity protocols and collaborate with federal agencies on risk assessment.
"You are seeing deals announced every day across all the spectrums - M&A, strategy, financing," Moynihan said.
The dual challenge of leveraging Mythos for defensive purposes while containing its potential for harm has prompted banks to invest heavily in cybersecurity protocols. Industry leaders are particularly worried that the same vulnerability-scanning power that makes Mythos valuable for red-teaming could be turned against financial infrastructure if the model ever leaks or is misused by insiders. This concern underpins the restricted access that Anthropic has granted only to a handful of trusted institutions.
Moynihan emphasized that the industry is working with outside partners too. "We are making sure the third parties are doing the work and working cooperatively with the team," he said.
Recent advances in AI have prompted both the financial sector and the federal government to collaborate on evaluating possible risks. This collaboration includes regular threat assessments and coordinated responses to emerging vulnerabilities. The industry is also exploring new standards for AI governance to ensure that powerful tools are used responsibly.
Anthropic markets Mythos as an advanced red-teaming instrument intended to assist companies in detecting security gaps prior to exploitation by bad actors. However, the financial industry's top executives have expressed concern that the same capabilities could be weaponized if the model falls into the wrong hands.
Bank of America's involvement in the testing group reflects a proactive approach to understanding the technology's implications. Financial institutions are also working to develop standardized protocols for sharing threat intelligence. Moynihan noted that the industry is coordinating with third-party vendors to ensure that security measures keep pace with the evolving capabilities of AI tools. This cooperative approach aims to create a unified defense against potential misuse of advanced models like Mythos.
Join Market Briefs, our free daily newsletter, for a quick daily rundown of the markets
