NextFin News - Sam Sidhu, the chief executive of Customers Bank, spent nearly thirty minutes of his first-quarter earnings call on Friday as a silent observer while an artificial intelligence clone of his own voice delivered prepared remarks to analysts. The revelation, which Sidhu shared toward the end of the call, served as a theatrical prelude to a more substantive shift: the $25.9 billion asset lender has signed a multiyear partnership with OpenAI to embed engineers directly into its operations.
The deal marks a significant escalation in the race to integrate generative AI into the plumbing of American finance. Under the terms of the agreement, OpenAI will assist the Pennsylvania-based bank in automating high-friction processes such as commercial lending and client onboarding. Sidhu told CNBC that the initiative is designed to lower the bank’s efficiency ratio—a key measure of profitability where lower is better—from approximately 49% to the low 40s. If successful, the bank expects to see a meaningful boost in returns starting in 2027.
The partnership is structured as a "symbiotic" co-development project. While Customers Bank gains a digital workforce capable of slashing loan processing times from over a month to just seven days, OpenAI gains a laboratory to refine "agentic" workflows—AI systems that can execute complex tasks rather than just generating text. These tools could eventually be marketed to other financial institutions, positioning Customers Bank as an early architect of OpenAI’s enterprise finance suite. Denise Dresser, OpenAI’s chief revenue officer, stated that the collaboration aims to set a "new standard for regional banking" by creating a more intelligent operating model.
However, the aggressive pivot toward AI-led automation is not without its skeptics. While Sidhu frames the move as a way to scale growth without a linear increase in headcount, the reliance on "AI agents" for underwriting and legal negotiations introduces new layers of operational and regulatory risk. Regional banks are already under intense scrutiny regarding credit quality; delegating the nuances of commercial lending to algorithms may invite pushback from regulators concerned about "black box" decision-making and the potential for systemic errors in automated document collection.
Market reaction to the bank's broader quarterly performance and its AI ambitions has been cautious. Shares of Customers Bancorp (CUBI) ended the previous trading session on Friday at $74.38, a decline of 3.99% as investors weighed the ambitious efficiency targets against the immediate costs of technological transformation. The bank’s strategy hinges on the assumption that AI can handle the complexity of commercial clients—whose accounts often take over a day to open—and collapse that timeframe to under 20 minutes without compromising compliance standards.
The move by Customers Bank reflects a broader trend where mid-sized lenders, often squeezed by the massive tech budgets of JPMorgan Chase and Bank of America, are turning to third-party AI giants to level the playing field. By embedding OpenAI engineers, Sidhu is betting that a regional player can leapfrog larger competitors in "agentic" banking. Whether this digital workforce can navigate the idiosyncratic risks of small-business lending as effectively as human bankers remains the central question for the bank's shareholders over the next eighteen months.
Explore more exclusive insights at nextfin.ai.
