🔒Liberty Bank makes a big play for AI adoption

Until recently, a Liberty Bank banker reviewing a small commercial real estate loan application would spend up to six hours gathering property data, analyzing financials and preparing an underwriting package before the deal could move to the next step. Today, artificial intelligence completes most of that work in about 15 minutes. While senior credit officers still make underwriting decisions, customers receive responses much faster and bankers spend less time on repetitive analysis. The dramatic reduction in processing time illustrates how AI is beginning to reshape banking at the 201-year-old, Middletown-based mutual lender. After seeing those early results, Liberty Bank launched an AI Center of Excellence in May. The new unit is intended to help identify where AI can improve the bank’s operations and oversee its implementation. Liberty, which has nearly $9 billion in assets, expects to invest up to $3 million over the next 12 to 18 months in AI initiatives, according to President and CEO David Glidden. “AI is not coming. AI is here and we need to get on board and we need to learn,” Glidden said. “You know, we need to use it and get efficiencies out of it. But you’re not going to do that unless you start using it.”

Glidden is already convinced of AI’s potential. The challenge, he said, is ensuring the technology is deployed with appropriate safeguards while keeping bankers — not machines — responsible for key decisions. Many Liberty employees already use tools such as ChatGPT and Microsoft Copilot to draft emails, summarize documents and conduct research. The AI Center of Excellence is intended to move beyond those basic tasks by identifying, testing and deploying artificial intelligence applications across the bank while ensuring new technology meets security and regulatory requirements, Glidden said.

Growing acceptance

Banks are increasingly adopting AI tools to improve internal operations, according to a 2025 survey by consulting firm EY-Parthenon. The survey found that 61% of responding banks reported significant impacts from internal use of generative AI, while 89% expected major benefits within two years. The report also identified significant hurdles, including data quality, regulatory compliance and implementation challenges. Even so, 38% of respondents said they expect to automate key business functions within five years, up from less than 10% in 2023. Liberty Bank’s AI initiative relies primarily on existing employees and is headquartered in the bank’s downtown Middletown Innovation Center. Its primary outside partner is Miami-based Flare AI, a startup led by a former Liberty executive. David Mitchell spent six years as Liberty’s chief digital officer, where he helped launch Owners Bank, the mutual lender’s nationwide digital banking platform for small businesses. He left Liberty in June to become president and chief strategy officer at Flare AI. Glidden said Liberty, like many banks and credit unions, is regularly approached by companies offering AI products. He said the bank wanted a trusted partner to help evaluate and implement the technology. “They’ll be one of our go-to people,” Glidden said. “When we find a use case, we can go to Flare, who we know, who we can trust.” Founded in 2024, Flare develops AI applications for banks and credit unions. Mitchell said the 55-employee company works with about a dozen financial institutions ranging from a $250 million community bank to one with roughly $3 trillion in assets. Mitchell’s path to Flare began in January when company founder and serial technology entrepreneur Scott Killoh asked what AI challenges Liberty was trying to solve. At that point, Mitchell said his digital team was spending nearly half its time manually testing software updates before they could be deployed across the bank, and he was looking for a way to streamline the process. Mitchell expected any solution would take some time. “I said to him, ‘It’ll take you six months to build this,’” Mitchell recalled. “And it took (Killoh) two weeks.” The resulting application, called Deploy IQ, uses AI to automate much of the software testing process. Mitchell said it now handles about 90% of that work. Deploy IQ is one of several AI applications Flare has developed for banks. Another one is Policy IQ, a tool that helps institutions update internal policies in response to changing regulatory guidance. Liberty is also using a tool called Lead IQ to help identify commercial lending opportunities for its recently opened Boston loan production office. The AI application analyzes market conditions and competitor activity to identify potential customers. Another application was developed after Glidden asked for a way to identify customers who might be seeking a new banking relationship following the pending merger of Webster Bank and Banco Santander. The tool analyzes branch locations, deposit concentrations and commercial lending data to pinpoint markets with the greatest opportunity. Mitchell said many community banks remain in the early stages of AI adoption because they lack the financial resources to build their own AI teams. That forces them to work with outside consultants. “This industry is undergoing a once-in-a-generation shift with AI,” Mitchell said. “There was when the internet was created and there’s this. It’s an AI tsunami right now. And community banks, as usual, are just far behind. The larger banks can afford developers and prompt engineers and a whole AI suite of people.”

A tool, not a replacement

While Glidden said he wants AI used broadly, he also wants to ensure its deployment is carefully vetted. To support that goal, Liberty created a formal governance process for evaluating new AI applications. The AI Center of Excellence is overseen by Chief Operating Officer Jeremy Miller, Chief Financial Officer Paul Young and Chief Risk Officer Troy Damboise. It is co-chaired by David Hadd, who leads the bank’s business transformation office, and Michael Stevens, Liberty’s data architect. Business unit leaders have been asked to identify repetitive, labor-intensive processes that could benefit from AI. Before any new application is adopted, the committee reviews cybersecurity, data privacy, regulatory compliance, vendor stability and how employees will interact with the technology. Glidden said Liberty views AI as a tool to make employees more efficient, not replace them. Bank employees will continue reviewing AI-generated work and making final decisions, he said. “So, one of our main goals here is to really look into what I call ethical AI,” Glidden said.