McKinsey predicts that while agent AI could reduce banks’ unit costs by 15 to 20%, it also threatens to erode up to $170 billion in global profit pools by 2030 if banks fail to adapt their business models.
Getty
While the public’s imagination has been captured by the conversational abilities of chatbots, a new report from McKinsey & Company suggests that the global banking industry is quietly approaching a much more profound transformation: the age of agentic artificial intelligence.
According to the report, the banking sector is moving from a period of widespread experimentation to a paradigm shift defined by autonomous agents, systems capable of programming, executing multi-step workflows and using tools with minimal human intervention. This transition represents the transition from “hype” to “precision”. It is no longer about inventing a machine that can write a poem. It’s about the utility of a system that can autonomously reconcile a ledger or transfer a mortgage.
However, this technological leap comes with a stern caveat. McKinsey predicts that while agent AI could reduce banks’ unit costs by 15 to 20%, it also threatens to erode up to $170 billion in global profit pools by 2030 if banks fail to adapt their business models.
For more like this on Forbes, What is Agentic AI and what will it mean for financial services?
To understand how financial institutions are bridging the gap between theoretical potential and production-level growth, I spoke with Jonathan Pelosi, Head of Financial Services at Anthropic, Scott Mullins, Managing Director of Financial Services at Amazon Web Services, and Steve Suarez, CEO of HorizonX, Senior Advisor at Globalnovation, GBCF and McKadsey.
Video: Jonathan Pelosi, Head of Financial Services at Anthropic
The 2026 Trust Horizon
For years, the adoption of artificial intelligence in banking has been limited by a trust gap. In a regulated industry, a fact-based model is a serious liability. Pelosi argues that this gap is closing rapidly due to the evolution of assessment frameworks.
“A year ago, when [researchers] checked, there might be 8 out of 10 facts that were correct,” Pelosi said. “Now you get like 99 out of 100.”
Pelosi identifies 2026 as the year the industry reaches a psychological and statistical tipping point. He draws a parallel with the adoption of autonomous vehicles. Just as passengers need data to trust a driverless car, bankers need data to trust an agent.
“An accuracy of 80% to 99% is impressive, unless you’re a bank,” Suarez said, adding, “With a 1% error a system is still misreporting 100 balances out of 10,000. Artificial intelligence in finance should aim for near-zero errors.”
Video: Scott Mullins, Managing Director of Financial Services at Amazon Web Services
Beyond AI tourism
As the technology matures, the industry’s approach to implementation matures with it. Mullins observes that banks are moving away from AI tourism, running pilots just to claim innovation.
“If what you’re trying to achieve is just ‘I want to do an AI experiment,’ that’s not really a real business outcome,” Mullins said. “What people see the most value in is having a very specific business outcome in mind.”
This shift from “wow” to “how” is driving banks toward what Pelosi calls “unsexy stuff.” The most impressive immediate use cases are not fancy chatbots, but deep functional improvements in the media and back office.
For more like this on Forbes, Legacy banks quietly building the future of finance.
The ‘Unsexy’ Revolution
One of the most critical applications for agent AI is to modernize aging industry infrastructure. Many financial institutions still rely on COBOL-based systems written decades ago.
“It turns out that these institutions are built on 40, 30-year-old legacy code that frankly people don’t even know how to code anymore,” Pelosi said.
He noted that Anthropic’s models are now successfully modernizing this legacy code, effectively reading millions of lines of archaic programming and recreating it in modern languages.
Likewise, compliance workflows such as KYC are moving from human-heavy processes to agent-driven automation. Mullins points to compliance reporting and risk management as areas where agents can significantly reduce manual intervention while improving accuracy.
However, integrating these agents requires navigating the reality of “fixing a flight while flying,” as Mullins described it. Banks cannot shut down core systems to upgrade them. they need to integrate AI agents into live, mission-critical environments.
The Disruption Threat: The Shopping Agent
While banks focus on internal efficiency, the McKinsey report highlights a major external threat: the rise of the purchasing agent.
Historically, banks have benefited from customer inactivity. It was simply too difficult for consumers to constantly switch accounts to find the best return. Agentic AI is ready to remove that friction. McKinsey predicts that consumer-facing AI agents will soon be able to autonomously monitor interest rates and move deposits to market-leading accounts.
If only 5 to 10 percent of checking balances were transferred to higher-yielding accounts originated by these agents, the industry’s deposit earnings could drop by 20 percent or more. This trend forces banks to compete not only with other banks, but also with the algorithms that manage their customers’ financial lives.
Governance: The Human-in-the-Loop
To address some of the risks, both Pelosi and Mullins emphasize the necessity of human governance in the loop. The goal is not to replace the banker, but to put the AI agent between layers of human supervision.
“You still have the added benefit that, while he can do 80-90 percent of the heavy lifting, people are still very much on board to make sure the checks and balances are in place,” Pelosi said.
Mullins advises CIOs to take a “golf bag” approach to this technology, using different models for different tasks rather than relying on a single vendor. This allows banks to choose the most secure and accurate tools for specific workflows, ensuring that governance evolves alongside technology.
Basics for bank executives
1. Aim for ‘Unsexy’ for high impact
Stop chasing innovation. Direct investment in “unsexy” back office bottlenecks such as legacy code modernization and automated compliance reporting. These areas offer the clearest path to the 15 to 20 percent cost reductions that McKinsey predicts.
2. Prepare for Algorithmic Competition
Recognize that customer inactivity ends. As purchasing agents begin to automate the change, banks must move from broad segmentation to a “department of one.” Use internal AI to proactively deliver hyper-personalized value to customers before an external agent moves their money elsewhere.
3. Operation of the “Sandwich” Governance.
Do not deploy agents without upgrading oversight processes. Implement a workflow where humans define goals and validate results, while agents handle execution. As Mullins warned, simply putting an agent into a workflow without accommodating human oversight is a recipe for failure.
