Stephanie Ferris, CEO and President, FIS
LUKE HEZEKIAH WALTER
On Monday at FIS Emerald in Orlando, fintech giant Fidelity National Information Services (FIS) announced a strategic partnership with Anthropic to bring agentic AI into banking, beginning with a Financial Crimes AI Agent designed to compress anti-money-laundering investigations from hours to minutes.
BMO and Amalgamated Bank will be among the first institutions deploying the system, with broader availability planned for the second half of 2026, according to FIS. On the surface, it’s a product announcement. In reality, it’s a much bigger signal about where AI in financial services is heading.
For the last two years, much of the AI conversation in fintech has centered around copilots, productivity tools, and customer-facing chat experiences. But financial services has always had a harder underlying problem to solve: trust.
Not just whether AI can generate answers, but whether it can operate safely inside one of the most regulated, high-stakes industries in the world.
Banking is quickly revealing that the hard part is not building the model, but deploying AI inside regulated systems where decisions must be explainable, auditable, and trusted every time.
That is the opening FIS, the financial technology company powering nearly 12% of the global economy, sees unfolding right now.
And it is why Anthropic partnered with them.
“Anthropic came to us because of our scale and the amount of data that we have,” FIS CEO and President Stephanie Ferris told me in an interview ahead of the announcement.
But the real moat, she argued, goes deeper than scale.
“Our moat is the 50 years of deep regulatory and compliance experience,” Ferris said. “If you take a product from us today, you know that it is absolutely built, completely compliant with whatever regulatory compliance you have.”
That distinction matters because it reframes the emerging power dynamic between frontier AI companies and financial infrastructure providers.
Anthropic brings Claude’s reasoning capabilities. FIS controls the environment where those systems are deployed: the infrastructure, governance, compliance architecture, and bank relationships required to operate AI safely at scale.
The AI company may provide the reasoning engine, but the financial infrastructure company governs where that reasoning operates, how it is audited, and whether it can ultimately be trusted inside the banking system.
Increasingly, that governance layer may become the more defensible business.
That operational complexity is precisely why Anthropic embedded its Applied AI and forward-deployed engineering teams directly with FIS.
“FIS brings decades of trusted relationships with financial institutions, deep regulatory knowledge, and the transaction data that makes an AI agent useful in practice,” Jonathan Pelosi, Anthropic’s Head of Financial Services, said in a statement. “They needed a model that could reason through complex investigations accurately, explain its work, and operate safely inside regulated workflows.”
The timing is notable beyond FIS alone.
In a separate announcement on Tuesday, Anthropic signaled a broader push into financial services, releasing 10 new AI agents designed for institutional workflows including building pitchbooks, reviewing earnings, drafting credit memos, and auditing financial statements.
But while many AI companies are approaching finance through productivity tools, FIS and Anthropic are positioning this partnership deeper inside the operational infrastructure of banking itself: compliance, fraud, governance, and transaction-level decisioning.
Or as Ferris put it more bluntly: “The hard part is, is it compliant? Is it correct every time, with every regulation, with every compliance? Can we audit it? Is it traceable?”
The partnership reveals a larger shift underway. The future for AI in banking is increasingly becoming a battle over who controls the orchestration layer: the data, regulatory workflows, client relationships, and trust required to deploy AI safely at scale.
Why Financial Crimes Came First
FIS is starting with financial crimes for a reason. The United Nations estimates that roughly $2 trillion in illicit funds moves through the global financial system every year. Banks collectively spend tens of billions annually on anti-money-laundering operations, much of it tied up in manual investigation work, fragmented systems, and compliance processes that remain painfully labor intensive.
The company’s new Financial Crimes AI Agent is designed to automatically assemble evidence across a bank’s core systems, evaluate activity against known typologies, and surface the highest-risk cases for investigator review, according to FIS.
The use case itself says a lot about where enterprise AI is maturing.
Consumer-facing AI demos attract headlines. But some of the largest immediate opportunities inside banking are operational: fraud investigations, compliance reviews, disputes, onboarding, servicing, and the thousands of manual workflows institutions still run across fragmented systems.
U.S. financial institutions spend between $35 billion and $40 billion annually on anti-money-laundering operations alone, according to FIS.
“We spend so much money as an industry on financial crimes, and the majority of that work is pulling things off systems and recalibrating data,” Ferris said. “It’s not necessarily the decision-making that the human agent makes.”
The company says early testing has already reduced manual process work by as much as 90%.
But the larger implication is strategic.
Financial crimes represents a controlled environment where AI must prove itself under the strictest conditions: auditability, human oversight, regulatory scrutiny, and direct financial risk.
If AI can survive there, banks may begin trusting it everywhere else.
Compliance Becomes The Competitive Advantage
For years, fintech treated compliance as friction. Now AI may turn it into infrastructure.
That shift is visible throughout FIS’s strategy.
Anthropic’s Applied AI and engineering teams are embedded directly with FIS to co-build the Financial Crimes AI Agent, while FIS maintains control of the governance, deployment, and client infrastructure underneath it.
Client data remains inside FIS-controlled infrastructure at all times, according to the company.
That point is not incidental.
It addresses one of the banking industry’s deepest concerns around enterprise AI adoption: losing control of customer data, regulatory accountability, or operational oversight to outside AI providers.
In other words, the partnership structure itself reflects the market’s emerging logic.
This also builds directly on a broader thesis Ferris has been articulating over the last several months around what she calls “orchestrated intelligence.”
Earlier this year, after FIS closed its $13.5 billion acquisition of Global Payments’ Issuer Solutions business, the company launched an “agentic commerce” offering for banks designed to help financial institutions safely manage AI-initiated transactions across payment networks.
At the time, Ferris argued that banks risked disintermediation if AI agents began shopping, negotiating, and transacting on consumers’ behalf without bank-controlled trust systems layered underneath.
Now FIS is building the operational infrastructure inside the bank itself.
Together, the moves point toward a much larger positioning strategy.
FIS is not trying to become the AI model company. It is positioning itself as the control layer between frontier AI and the financial system.
And that may prove to be one of the more important positions in fintech.
Why The Power Dynamic May Be Changing
For years, the fear across banking and fintech was that big tech would eventually abstract away the financial institutions themselves.
But AI is reshaping that equation.
Because while frontier AI companies can build extraordinary reasoning systems, regulated industries introduce constraints that are difficult to replicate quickly: compliance expertise, auditability, transaction infrastructure, risk systems, supervisory relationships, and trusted access to financial data at scale.
That is why this partnership matters beyond one product launch.
Anthropic is not bypassing financial infrastructure players.
It is embedding inside them.
“You’re not going to innovate around us,” Ferris told me. “If you’re going to innovate around us, you wouldn’t have Anthropic strategically partnering with us.”
That is a striking statement from the CEO of a legacy fintech infrastructure company, because it reflects a broader market reality emerging across financial services: AI may accelerate the importance of infrastructure incumbents rather than weaken them.
Especially the ones capable of translating frontier models into regulated, bank-grade systems that can actually operate at scale.
Because in financial services, intelligence alone is not enough.
The system, and the humans behind it, have to trust the machine making the decision.