Banking giants aren’t slowing their AI investments as they pursue efficiency and productivity gains from the technology’s implementation. However, the latest frontier model developments are causing a cybersecurity stir among financial institutions.
Firms including JPMorgan Chase, Morgan Stanley, Goldman Sachs, BNY and others reported earnings results for the first quarter of 2026 this week. While executives focused on AI wins and the technology’s potential in the industry, several fielded concerns about frontier AI models and their implications for cybersecurity.
Anthropic earlier this month said its Claude Mythos Preview frontier model could reshape cybersecurity as it surpasses human capabilities of finding and exploiting software vulnerabilities. The model provider claims Mythos Preview has already discovered thousands of serious vulnerabilities across major web browsers and operating systems. Anthropic formed Project Glasswing with partners, including JPMorgan Chase, as a result. Launch partners will be using Mythos Preview in their defensive cybersecurity work, according to Anthropic.
“JPMorgan is very well protected,” CEO Jamie Dimon said during the company’s Q1 2026 earnings call Tuesday. “We spend a lot of money, we’ve got top experts, we’re in constant contact with the government. We’re constantly updating things, but AI has made it worse, it’s made it harder. Of course, we read about Mythos, which we’re testing now and looking at it and it does create additional vulnerabilities.”
Morgan Stanley CEO Ted Pick also addressed cybersecurity concerns around AI frontier models during the company’s earnings call Wednesday. The company is working with Mythos Preview to improve its cybersecurity infrastructure, an area the firm has long invested in, Pick said.
“AI is our friend, OK?” he said. “It is just the latest generation of technology that is going to be part of the ecosystem.”
For nearly all banking executives, data security, privacy and risk will be the top factors affecting AI strategy over the next six months, with 80% of executives including cyber and data security in their AI budgets, according to KPMG’s AI Quarterly Pulse Survey.
Projected AI spend for banks over the next 12 months reached $177 million in Q1 2026, a 33% increase from the previous quarter, the survey found. Banking institutions’ AI investments will continue to grow, moving from isolated use cases to enterprise-wide integration, said Chris Long, KPMG U.S. advisory leader in financial services.
“The priority now is scaling AI capabilities across the organization while strengthening governance, monitoring and controls to detect misuse, prepare for bad actors, and align risk, compliance, and business functions to support responsible growth,” he said in an emailed statement.
AI gains in banking
Despite AI risk concerns, banks are capitalizing on the technology’s benefits and pouring resources into business transformation.
Goldman Sachs is leaning on AI to accelerate growth and efficiency within the company’s business operations. The bank plans to double down on infrastructure as it builds out its One Goldman Sachs 3.0 AI operating model, including accelerating investments in cloud migration and data accuracy, CFO Denis Coleman said.
“These investments are critical to optimizing the deployment of AI solutions across the firm, which will allow us to unlock greater productivity and efficiency opportunities over time,” Coleman said during the Monday earnings call.
Morgan Stanley is also zeroed in on AI gains, including moving beyond small wins such as automating routine operational functions to transformational business overhauls leading to productivity boosts, Pick said.
“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,” he said.
BNY has focused on broad AI adoption, developing more than 200 AI products such as agentic digital employees operating alongside humans, CEO Robin Vince said during the Thursday earnings call.
The banking giant plans to increase AI adoption in 2026 as it embeds the technology in workflows across the company and focuses on using AI to enhance end-to-end processes, he said, adding that the technology is already contributing to the bank’s financial performance.
“AI is helping us increase the pace at which we innovate our technology, accelerate onboarding, improve client service and streamline processes,” Vince said.
Security is also top of mind for the bank. BNY is running the Mythos Preview model and is part of OpenAI’s Trusted Access for Cyber program, which the model provider expanded in preparation for future, more capable AI models.
“What AI really is now is a superpower,” Vince said. “It can be used for good and it can be used for evil. We’re pulling the superpower into our environment to use for good in order to be able to defend ourselves. We view this as an entirely predictable evolution of the technology.”
End-to-end process improvement and cybersecurity are also on Citigroup’s radar. The financial services firm is using AI for operational efficiency gains by simplifying complex and manually-intensive processes, as well as cybersecurity defenses, said CEO Jane Fraser during the company’s earnings call Tuesday.
Citi’s enterprisewide approach to AI also focuses on revenue generation and workforce implications, Fraser said.
“We are methodically deploying AI at scale across the firm to drive revenues and process improvements, enhance client experiences and strengthen our defensive capabilities,” she said.
