DR Bank is opting for a decentralized approach to deploy AI. 

AI decisions for the bank are made by an AI council made up of volunteers from every corner of the bank, Chief Executive Jason Hardgrave told FinAI News.  

 “I want people who are excited,” Hardgrave said. “They were all using ChatGPT or Claude on the side anyway.” 

The council is intentionally title-agnostic — one seat is held by a junior staff accountant, another by the bank’s chief compliance officer, he said, adding that department heads are not automatically included. The bank did not disclose the number of employees in the AI council.

The goal, Hardgrave said, is self-direction: Each member identifies what to automate within their own function and pitches it to the council. Once approved, IT teams are deployed to make the process more efficient.  

The decentralized approach allows the bank to identify bottlenecks that it didn’t know existed, he said.  

AI deployment 

The bank has deployed more than 75 automations and six AI agents across its finance, compliance and operations functions over the past two years, Hardgrave said.  

He draws a line between automations and agents. 

Automations handle predictable, repeatable tasks — reconciling transaction files in which the output must be identical. Agents handle variable queries, such as a call center tool that answers employee questions about bank policy and cites the relevant document for verification. 

“In a lot of cases, we’re not using AI to do the work,” he said. “We’re using AI to code the work, to create the automation where we’ve never been able to do it fast enough.” 

The bank has redeployed staff freed up by automation into more skilled roles, Hardgrave said. 

Ninety percent of the bank’s AI use is back-office, Hardgrave said, adding that since DR Bank is a sponsor bank, there is little need for AI in customer facing roles. The bank has deployed a multi-model strategy and uses Anthropic‘s Claude and Google‘s Gemini for its AI needs among other AI tools.

AI agents at DRB 

Of the six AI agents that have been deployed, three have names worth knowing. 

Loan Wolf handles loan quality control, sampling every loan file and flagging those that require human investigation, a shift from the previous approach of manually reviewing a 10-file sample, Hardgrave said.  

ERMie, short for enterprise risk management, reviews third-party due diligence documents, including SOC1 and SOC2 reports, against a predefined compliance checklist, a process that previously took hours, he added.  

ERMie reviews reports from fintech partners as part of third-party due diligence, essentially checking whether a partner’s controls meet the bank’s requirements before flagging exceptions for a human reviewer, he explained. 

The third agent, Comply AI, runs regulatory tests including TILA [Truth in Lending Act] and Reg B. 

The Darien, Conn.-based bank has also deployed AI agents for call center operations and an internal chatbot within the bank, Hardgrave said. 

AI and sponsor banks 

The AIcouncil’s work sits inside a broader growth story at DR Bank. 

The $860 million bank has grown its balance sheet roughly 20% year over year in 2026 so far, he said, adding that the bank partners with three to four fintechs a year. DR Bank acts as a sponsor bank for many fintechs like Doc2Doc Lending and Upstart. 

The sponsor banking model, which faced regulatory headwinds a few years ago, has benefited from fintechs pursuing their own bank charters, Hardgrave said.  

“If you’re going to give a fintech a charter, you can’t say anything bad about a bank partnering with a fintech.” 

FinAi Lending Summit, set for Oct. 7-8 in Las Vegas, will include speakers from Fifth Third and Capital One as well as a fireside chat with U.S. Bank Senior Vice President – Lending Operations Group Manager Suzanne Rathbun. To learn more about the 2026 event and register, visit here.