Updated July 2026: This article has been refreshed to clarify what a national trust bank charter would—and would not—allow, tighten the regulatory context around AI agents handling money, and remove unsupported claims about product capabilities that have not yet been publicly demonstrated.
Why Building Banks for AI Agents Could Reshape Financial Services
Catena Labs has raised $30 million to build regulated financial infrastructure for AI agents—autonomous software systems that can execute tasks, make decisions, and potentially initiate financial transactions on behalf of people or businesses. According to Bankless, the company is also pursuing a national trust bank charter, signaling that it wants to build inside the regulatory perimeter rather than simply attach AI tools to existing banking APIs.
The core bet is straightforward: if AI agents are going to pay invoices, move funds, manage subscriptions, optimize treasury workflows, or execute business operations, they will need financial rails designed for delegation, permissions, auditability, and liability. Legacy banking systems were built around human account holders, human authentication, and human approvals. AI agents complicate that model because they may act continuously, interact with multiple systems, and make decisions within parameters set by users or companies.
That does not mean AI agents should be able to move money without oversight. In fact, Catena’s thesis appears to be the opposite: agentic finance will require more explicit governance than ordinary online banking. The opportunity is to create infrastructure where agents can act only within defined limits, every action is logged, and compliance teams can understand who authorized what, when, and why.
Catena Labs’ $30M Funding: What the Numbers Reveal About Investor Confidence
A $30 million raise is a meaningful vote of confidence in a category that is still early: regulated financial infrastructure for AI-native activity. The funding suggests investors see agentic finance as more than a speculative use case. If AI agents become standard in enterprise software, e-commerce, treasury, and personal productivity, they will need trusted ways to interact with money.
Catena is not positioning itself as another budgeting app or chatbot layered onto a bank account. Its emphasis is on “governed infrastructure”—systems that can define permissions, enforce transaction policies, verify authority, produce audit trails, and support regulatory review. That distinction matters. The technical problem is not simply whether an AI model can decide to pay a bill; the harder question is whether a regulated financial system can verify that the agent had permission to do so and that the transaction complied with applicable rules.
The size of the round also reflects a broader investor shift. AI infrastructure funding has moved beyond model labs and developer tools into vertical layers: legal AI, healthcare AI, enterprise agents, and now financial execution. In finance, however, speed alone is not enough. The winners will likely be companies that can combine automation with compliance, identity, fraud controls, and operational resilience.
Navigating Regulatory Hurdles: The Significance of Filing for a National Trust Bank Charter
Catena’s reported application for a national trust bank charter with the Office of the Comptroller of the Currency is one of the most important parts of the story. A national trust bank charter is not the same as becoming a full-service FDIC-insured commercial bank. It generally supports limited-purpose banking activities such as custody, fiduciary services, and related financial functions, depending on the approved business plan and regulatory conditions.
That distinction is important because early coverage of AI banking can easily overstate what a trust charter allows. Approval would not automatically let Catena offer ordinary checking accounts, make loans, or operate like a traditional retail bank. But it could give the company a federally supervised framework for holding or administering assets, building compliance controls, and operating infrastructure that connects AI agents to financial workflows.
For AI agents, that regulatory posture could be a competitive advantage. Regulators are unlikely to be comfortable with autonomous systems moving money at scale through lightly supervised workarounds. A chartered structure would force Catena to address governance, risk management, cybersecurity, anti-money laundering controls, third-party oversight, and operational continuity from the start.
The charter path is also slow and uncertain. OCC review can be lengthy, and approval—if it comes—may include conditions, capital requirements, activity limits, or supervisory expectations. Catena’s real test is not just whether it can build agentic banking software, but whether it can satisfy regulators that the model is safe, auditable, and legally accountable.
Diverse Stakeholder Perspectives on AI-Driven Banking Infrastructure
Fintech innovators see agentic finance as a new platform layer. If agents become the interface through which people and businesses manage financial tasks, then the underlying permissioning and transaction infrastructure becomes extremely valuable. Developers need a way to let agents interact with money without building compliance systems from scratch.
Regulators will focus on a different set of questions. Who is responsible when an AI agent initiates a mistaken payment? How are customers authenticated? Can users revoke permissions instantly? What happens if an agent is manipulated by prompt injection, phishing, compromised credentials, or malicious third-party data? How are suspicious transactions monitored when the initiating actor is software?
Traditional banks may view Catena as both a threat and a potential partner. Banks already have licenses, compliance teams, customer relationships, and payment connectivity. But many are not architected for a world where software agents act as delegated financial operators. If Catena proves the model, banks may seek to partner with or replicate similar infrastructure.
For consumers and businesses, the promise is convenience with controls. A small business could authorize an AI agent to pay approved vendors up to a limit, reconcile invoices, or move idle cash according to policy. A consumer could delegate subscription management or bill payment. But the risks are real: automation can amplify mistakes, and financial fraud may become more sophisticated when attackers target agents rather than humans.
From Traditional Banking to AI-Powered Agents: A Historical Shift
Banking technology has evolved from branch ledgers to ATMs, online banking, mobile apps, open banking APIs, and embedded finance. Yet the central assumption has remained mostly unchanged: a human or legal entity owns the account and directly authorizes activity.
AI agents challenge that assumption. They are not legal persons, but they can act on behalf of legal persons. That creates a new design problem: financial systems must distinguish between the customer, the agent, the software provider, the model, the device, and the specific authorization behind each transaction.
Previous fintech waves often improved the user interface while relying on existing banks behind the scenes. Catena’s wager is that agentic finance requires deeper infrastructure, not just a better app. The key layer is not a chatbot that answers banking questions; it is the regulated control system that determines what an agent may do with money.
The historical analogy is less “AI replaces banks” and more “new transaction types require new rails.” Card networks, ACH, real-time payments, stablecoins, and open banking all emerged because existing systems did not fully satisfy new forms of commerce. AI agents may create another such shift if delegated financial action becomes common.
What AI-Enabled Banking Infrastructure Means for Financial Industry Stakeholders
For banks and fintechs, Catena’s approach raises a strategic question: should AI agents be treated as a feature, a customer channel, or a new class of delegated operator? The answer affects compliance architecture, product design, fraud controls, and partner strategy.
If Catena establishes a credible regulatory template, other financial institutions may need to support agent-specific permissions, transaction logs, risk scoring, and revocation tools. AI developers could benefit from standardized financial access layers instead of negotiating one-off integrations with banks and payment processors.
For enterprises, the potential upside is operational efficiency. Finance teams spend enormous time on repetitive workflows: approvals, payments, reconciliations, expense management, cash movement, and compliance documentation. Properly governed agents could reduce manual work while preserving oversight.
The downside is that financial automation failures can be costly. A hallucinated email is embarrassing; a hallucinated wire transfer is dangerous. Any platform serving AI agents must assume that models can be manipulated, credentials can be compromised, and users may misunderstand what they authorized. Strong controls are not optional—they are the product.
What We Know, What Remains Unclear, and What to Watch
What’s clear: Catena Labs has raised $30 million to pursue regulated financial infrastructure for AI agents, and it has reportedly applied for a national trust bank charter, according to Bankless. The company is betting that AI agents will need purpose-built financial rails with governance, auditability, and regulatory oversight.
What remains unclear: Catena has not publicly proven at scale how its platform will work in live financial environments. Key open questions include customer onboarding, agent identity, permission management, fraud liability, dispute handling, cybersecurity, and the exact scope of activities allowed if a charter is approved. The OCC timeline also remains uncertain.
What to watch: The most important milestone is whether Catena receives regulatory approval and under what conditions. After that, watch for pilot customers, banking or fintech partnerships, stablecoin or payment integrations, and detailed product disclosures. The first real-world stress test will come when AI agents begin moving meaningful sums under explicit policy controls and regulatory supervision.
If Catena succeeds, it could help define how money moves in an AI-agent economy. If it fails, the broader lesson may still stand: financial institutions cannot simply give autonomous agents access to legacy accounts and hope existing controls are enough.
Disclaimer: This MLXIO analysis is for informational and educational purposes only. It is not financial, investment, legal, tax, or professional advice. It does not provide buy, sell, hold, price-target, portfolio, or personalized recommendations. Verify information independently and consult qualified professionals before making decisions.
Why It Matters
- Catena Labs’ $30M raise highlights growing investor interest in regulated banking infrastructure for AI agents.
- A national trust bank charter could give Catena a supervised framework, though not the same powers as a full commercial bank.
- Agentic finance may reshape payments, compliance, fraud prevention, and delegated financial automation.
