BURBANK, Calif.—Financial institutions are racing to figure out where generative AI fits into their lending operations, and a new Celent study offers one of the clearest snapshots yet of how quickly the technology is taking hold—and where lenders expect real impact.
The report, commissioned by Zest AI, finds banks and credit unions planning sharp increases in GenAI spending next year and moving to integrate the tools into underwriting, automation, and workflow design at a pace that outstrips previous technology cycles.
The "Generative AI in Retail Lending" survey, conducted in August 2025, polled 106 U.S. banks, credit unions, and consumer finance companies and found that 83% of lenders plan to increase their consumer lending GenAI IT budgets in 2026, with 41% anticipating increases exceeding 5%.
Most significantly, two-thirds (67%) of lenders have already completed or will implement GenAI strategies by 2026 – a faster adoption rate than previous lending technology innovations, including AI/ML, online lending, mobile lending, and statistical regression-based automated underwriting.
The Race To Responsible GenAI
Faced with rising interest rates, intense regulatory scrutiny, and competition from digital-first lenders, financial institutions are turning to GenAI to enhance customer experience (21%) and reduce expenses and improve operational efficiency (20%). The Celent report revealed several notable insights about how lenders are approaching GenAI adoption:
- Consumer lending divisions are catching up to other divisions across retail banks: Only 6% of consumer lending divisions have a GenAI strategy today, but 36% are developing one for 2025 – putting them on track to close the gap with the broader retail bank, where 15% have already implemented GenAI and another 32% plan to do so by year-end
- Overcoming barriers: Security and risk concerns remain the biggest hurdle to GenAI adoption, with 19% citing them as their top concern and another 20% ranking them second. Competing priorities followed as the next major challenge, noted by 15% of respondents
- Broader AI ecosystem growth: Beyond GenAI, 58% of lenders have Robotic Process Automation in production, 39% use AI/ML for loan decisioning, and Agentic AI is on the rise – 8% are already using it and 28% are planning deployment
“Generative AI is being adopted faster in lending than any previous AI technology wave we’ve tracked,” said Craig Focardi, principal analyst at Celent. “What’s striking is how quickly financial institutions are moving from pilots to production. The urgency to compete on both efficiency and customer experience is turning GenAI from a curiosity into a core capability.”
The Competitive Opportunity
The Celent survey also highlighted a significant competitive opportunity, with 36% of lenders viewing themselves as early adopters of GenAI. Respondents identified three key areas where GenAI delivers the most competitive differentiation: employee copilots that serve as AI assistants for loan officers, processors, and underwriters to query policies and documents conversationally; compliance and regulatory reporting that automates documentation and flags potential violations; and business intelligence enhancements that provide advanced analytics and reporting capabilities.
From Experimentation To Transformation
Seventy percent of lenders are adopting or planning to adopt GenAI for analytics and reporting dashboards – a faster adoption rate than most consumer lending use cases. However, this rapid adoption is outpacing the development of AI legislation, creating potential compliance risks for institutions that haven't prioritized responsible AI development, Zest AI explained.
“It's crucial for lenders to choose tools built with transparency, fairness, and ethical practices from the ground up – qualities that will become regulatory requirements, not just best practices. Zest AI’s LuLu, the leading generative AI lending intelligence platform built for financial institutions, converts that adoption into measurable impact, enabling institutions to consolidate up to 50% of their reporting tools, reduce analysis time and staffing by 80%, and accelerate decision-making by as much as 90%. Together, these gains allow lenders to compete on speed and efficiency without sacrificing the transparency and compliance that regulators demand,” the company said.
The Takeaway
Zest AI concluded that Celent’s findings mark a decisive shift from experimentation to execution.
“The next frontier for lenders is about applying GenAI strategically across the business to drive measurable impact. Success will depend on solutions that fit seamlessly into existing systems, solve real operational challenges, and scale securely,” the company stated.
“We’re seeing a genuine inflection point in how lenders think about AI,” said Mike de Vere, CEO of Zest AI. “The conversation has moved beyond experimentation to practical, measurable transformation. What matters now is applying generative AI responsibly and in ways that truly support lenders, customers, and regulators alike.”
The complete Celent report, "Generative AI in Retail Lending," is available here.