The traditional hierarchy of banking technology has long been defined by a stark divide. On one side, the "megabanks"—the likes of JPMorgan Chase and Bank of America—possess the multi-billion-dollar R&D budgets and legions of software engineers required to build proprietary systems from the ground up. On the other side, midsize and regional banks have historically been "takers" of technology, relying on a fragmented ecosystem of third-party vendors to provide niche, purpose-built solutions for everything from loan origination to fraud detection.
However, a fundamental shift is underway. The emergence of agentic artificial intelligence—AI capable of autonomous reasoning and executing complex workflows—is upending the "build-versus-buy" calculus that has governed the sector for decades. Leading this charge is Valley Bank, a $64 billion-asset institution based in Morristown, New Jersey, which is increasingly opting to develop its own AI-driven capabilities rather than renewing legacy vendor contracts.
Main Facts: The Strategic Pivot at Valley Bank
Valley Bank’s Chief Operating Officer, Russ Barrett, recently detailed a significant transformation in the bank’s operational philosophy. For midsize banks, the standard operating procedure has been to acquire "best-of-breed" niche products to solve specific problems. While effective in the short term, this approach often leads to a "Frankenstein’s monster" of disconnected software, high licensing fees, and data silos.
According to Barrett, Valley Bank is already in the process of replacing three external vendor contracts by leveraging internal AI capabilities. These were not minor tools; they were software solutions focused on specific, critical workflows. The move signifies a growing confidence among midsize banks that they can now compete with—and in some cases, outpace—the engineering prowess of larger institutions by utilizing AI as an "exponential improver" of their internal engineering teams.
The core of this shift lies in "Agentic AI." Unlike standard generative AI, which might simply summarize a document or draft an email, agentic AI can act as a digital employee, navigating through different software environments, making decisions based on data, and completing multi-step processes. For a bank like Valley, this means the ability to create custom "agents" that handle internal efficiencies and relationship-bolstering tasks that were previously outsourced to expensive third-party platforms.
Chronology: From Core Conversion to AI Production
The path to Valley Bank’s current AI maturity was not an overnight achievement. It was the result of a multi-year strategic roadmap designed to modernize the bank’s digital foundations.
Phase 1: The Core Conversion and Cloud Integration
Before a bank can effectively deploy AI, its data must be accessible and clean. Valley Bank spent significant resources on a "core conversion"—a massive undertaking that involves replacing the central processing system that handles a bank’s most basic functions. By moving to a cloud-first strategy, Valley ensured that its data was no longer trapped in legacy on-premise servers. Barrett contends that this foundational work gave the bank a "head start," allowing it to integrate AI tools much faster than peers who are still struggling with fragmented data architectures.
Phase 2: Rapid Prototyping and Production
With the cloud infrastructure in place, Valley began partnering with AI firms to develop prospecting and fraud-detection capabilities. A key differentiator for Valley has been its speed to market. Barrett noted that the bank reached the production phase with these AI-powered tools significantly sooner than other similarly sized banks working with the same partners. This phase marked the transition from "talking about AI" to "realizing benefits from AI."
Phase 3: The Current State of "Token-Conscious" Scaling
In early 2024, as AI costs began to skyrocket across the corporate world, Valley Bank implemented a disciplined, tiered access model. Rather than a "free-for-all" approach, the bank now manages AI usage through a structured system of five employee tiers, ensuring that expensive "power tools" are reserved for those who can generate the highest return on investment (ROI).
Supporting Data: The Economics of the AI Shift
The financial and industrial context of this shift is underscored by recent market data and Valley’s own fiscal reporting.
1. Rising Tech Expenditures
In the first quarter of 2024, Valley Bank reported that its technology, furniture, and equipment expenses rose approximately 7% year-over-year, reaching $31.9 million. While this increase was largely driven by data processing fees and software licensing, it highlights the fiscal pressure that midsize banks face. By building AI tools internally to replace three external contracts, Valley aims to offset these rising costs and eventually bend the expense curve downward.
2. The D.A. Davidson Industry Survey
A recent survey by D.A. Davidson of 73 banks across various asset sizes confirms that Valley is part of a growing trend, albeit an early mover.
- 42% of bank respondents are currently providing individual AI user tools to staff.
- 35% have identified and implemented quantifiable, tangible use cases.
The survey noted that the emergence of "tangible use cases" over the past few months has been a notable shift from the theoretical discussions of 2023.
3. The "Tokenmaxxing" Warning
Valley’s cautious approach stands in contrast to the broader tech industry. For example, the rideshare giant Uber reportedly exhausted its entire 2026 AI coding budget in just four months. This phenomenon, often called "tokenmaxxing" (maximizing AI use without regard for cost), is a trap Valley is determined to avoid. By being "token-conscious"—tracking the unit of measure for AI usage—Valley ensures that every AI interaction has a clear "why" behind it.
4. Personnel Allocation
Of Valley’s roughly 3,600 employees, only about 80 are currently in the "open access" tier. These are the "power users" who have the latitude to build and experiment with custom AI agents. The rest of the staff uses pre-built, controlled AI agents that provide specific efficiencies without the risk of runaway costs.
Official Responses: Insights from COO Russ Barrett
Russ Barrett’s philosophy on AI is rooted in pragmatism rather than hype. He makes a sharp distinction between signing a contract and delivering a result.
"A lot of people talk about the first [identifying a use and signing a contract], and it’s a little bit more challenging to be able to sit there and stick to it to see exactly how to get more to the latter [getting a tool into production]," Barrett said. "We probably are not market-leading in the number of use cases, but we do feel our ability to execute and see it to fruition is definitely something."
Regarding the customer-facing side of the technology, Barrett emphasized a "not-creepy" approach. The goal is to use AI to be a "better relationship bank" by identifying leads and prospecting more effectively, but without making the customer feel like they are being monitored by a machine. "This is a journey," Barrett explained. "It is not, you’re going to go buy a tool, and all of a sudden your door is going to be kicked down with all the customers you’re able to generate, because you’re still operating with a very people-centric model."
Barrett also addressed the ROI of AI, noting that it isn’t always about immediate cost savings. "It’s a formula that allows us to really understand, how do we exponentially add more value to our credit people, to our salespeople, than what we would have today?"
Implications: A New Era for Midsize Banking
The shift at Valley Bank has profound implications for the future of the financial services industry, particularly regarding vendor relations and workforce management.
1. The Disruption of the Fintech Vendor Model
For years, fintech companies have thrived by selling "point solutions" to midsize banks. If banks like Valley can now build these solutions internally using agentic AI, the valuation and necessity of many niche fintech vendors may plummet. The "build-versus-buy" calculus is tilting toward "build" because AI acts as a force multiplier for a small team of internal engineers, allowing them to do the work that previously required an entire software company.
2. The Rise of the "Augmented Banker"
Valley Bank is not using AI to replace its 3,600 employees but to eliminate "low-quality, repeatable work." This suggests a future where the role of the banker shifts from data entry and manual prospecting to high-level advisory and relationship management. However, the bank’s lesson that AI cannot yet "codify" the behaviors of relationship building means that human soft skills will remain the primary differentiator in the market.
3. Fiscal Discipline as a Competitive Advantage
As demonstrated by the Uber example, AI can be a "black hole" for capital if not managed correctly. Valley’s tiered access and "token-conscious" strategy may become the industry standard for fiscal responsibility. Banks that fail to monitor their AI consumption may find themselves facing "bill shock" that erases any efficiency gains the technology provided.
4. Closing the Engineering Gap
Perhaps the most significant implication is the leveling of the playing field. If a $64 billion bank can develop proprietary tools that rival those of a $3 trillion bank, the competitive advantage of sheer size begins to diminish. Agentic AI allows midsize banks to be more agile, creating bespoke tools that fit their specific community-banking culture rather than being forced to adopt the rigid structures of "off-the-shelf" vendor software.
In conclusion, Valley Bank’s strategic pivot marks a turning point. The bank is no longer just a consumer of technology; it is becoming a creator. As agentic AI continues to evolve, the ability to execute and move tools into production will separate the winners from the laggards in the increasingly digital landscape of American banking.
