The Generative AI Revolution: Nearly Half of US Banks Embrace GenAI in 2025

US Banks Accelerate GenAI Adoption in 2025, Signaling Strategic Shift

The year 2025 marks a significant inflection point for the US banking sector, with nearly half of all institutions having fully rolled out generative AI (GenAI) capabilities. Data reveals that 47% of US banks have achieved full deployment of GenAI, a dramatic increase from a mere 10% in 2023. This rapid acceleration signifies a decisive move away from experimental pilot projects towards strategic, large-scale implementation across the industry. Financial institutions are increasingly recognizing that delaying GenAI adoption poses a substantial risk, potentially leading to competitive disadvantages in customer experience, operational efficiency, and the pace of innovation.

Investment Surges as Banks Embrace GenAI

The commitment to GenAI is underscored by a substantial increase in investment. A March survey by Capgemini found that over two-thirds (67%) of senior banking executives reported heightened investment in GenAI over the preceding year. This financial commitment is a key enabler of the widespread adoption observed in 2025, fueling the development and deployment of advanced AI solutions.

Consumer Trust and the Human Element in AI Interactions

While banks are rapidly integrating GenAI, consumer sentiment highlights the continued importance of human oversight. A June survey by Auriemma Group indicated that a significant majority of US debit card holders—56% for disputed transactions and 55% for customer service issues—deem human intervention to be very important. This suggests a nuanced approach is required, where GenAI enhances efficiency and capabilities, but human interaction remains critical for building trust and managing sensitive customer needs.

EY-Parthenon Report: Progress and Future Outlook

Further insights from the EY-Parthenon report on "GenAI in Retail and Commercial Banking" paint a detailed picture of the industry’s GenAI journey. In 2025, 77% of banks have either launched or soft-launched GenAI applications, a considerable rise from 61% in 2023. The impact is already being felt, with 61% of surveyed banks reporting substantial benefits from their current deployments. Looking ahead, an impressive 89% anticipate major transformative benefits within the next two years. The potential for full end-to-end automation is also a growing expectation, with 38% of respondents projecting this within five years, a significant jump from just 8% in 2023.

Diverse Use Cases and Front-Office Prioritization

GenAI applications are being deployed across various banking functions, with a balanced distribution across the front office (33%), middle office (35%), and back office (31%). However, front-office use cases, such as customer service and marketing, represent a larger proportion of applications already in production (43%). This strategic focus on customer-facing applications suggests a prioritization of areas with potentially higher and more immediate returns on investment (ROI), leveraging AI to enhance customer engagement and service delivery.

Balancing GenAI with Traditional Automation

While the excitement around GenAI and the emerging field of agentic AI is palpable, traditional machine learning (ML) and robotic process automation (RPA) remain foundational. Currently, GenAI and agentic AI constitute only 28% of automation use cases in development or implementation. This indicates a strategic approach by banks to select the most appropriate technology—whether GenAI, ML, RPA, or agentic AI—for specific tasks to maximize value and efficiency.

The Rise of Agentic AI and Implementation Hurdles

Agentic AI, characterized by its ability to make autonomous decisions and take actions, has rapidly gained recognition, with 99% of banking respondents aware of its potential. However, actual implementation is still in its early stages, with only 31% of banks having moved towards deployment. The path to widespread adoption is being navigated with caution, as challenges related to data, regulatory compliance, and execution have led to notable development and implementation failure rates. Overcoming these hurdles and integrating lessons learned from past initiatives are crucial for unlocking the full strategic value of GenAI and agentic AI.

Financial Impact and Investment Strategies

The financial implications of GenAI adoption are a significant driver for investment. Fifty-eight percent of banks anticipate a positive revenue uplift ranging from 6% to 20% from their GenAI applications. Furthermore, 79% of respondents expect even greater revenue uplifts in the next two years. This projected financial return continues to fuel investment, even as banks refine their strategies to mitigate implementation risks and failure rates.

Evolving Governance and Funding Models

In response to the complexities of AI integration, banks are increasingly adopting centralized governance models to streamline decision-making and enhance accountability. A substantial 75% of banks now have formal governance committees, with 60% empowering executive leadership teams with decision-making authority. Investment in GenAI is increasingly being channeled through IT and technology budgets (65%), marking a shift from broader corporate strategy funding. Banks are also leaning more heavily on external partnerships (57% of planned use cases), indicating a strategic move towards leveraging external expertise to accelerate deployment and overcome internal resource constraints.

Strategic Recommendations for Future Success

To effectively navigate the evolving landscape of AI in banking, institutions are advised to adopt a strategic framework focused on key recommendations:

• Early Stakeholder Engagement: Actively involve key internal stakeholders from the outset of GenAI planning to ensure alignment and minimize execution risks.
• Clear Technology Alignment: Develop strategies that clearly differentiate the application of GenAI, agentic AI, RPA, and traditional ML, ensuring the right technology is matched with specific use cases for maximum impact.
• Enhanced Data Governance: Address data-related barriers by improving data governance frameworks and prioritizing data quality initiatives.
• Prioritize Partnerships: Leverage external collaborations to rapidly scale capabilities and bridge internal skill and resource gaps.
• Strengthen Governance and Controls: Implement robust governance structures and compliance measures, particularly for high-risk and customer-facing GenAI applications.

By focusing on these strategic imperatives, banks can effectively harness the power of GenAI and agentic AI to drive innovation, optimize operations, and secure a competitive advantage in the rapidly transforming financial services landscape.

Survey Methodology

The insights presented are derived from a targeted survey conducted by EY-Parthenon in early 2025. The survey engaged 100 senior decision-makers across leading retail and commercial banking firms in the United States. Participants held strategic and operational leadership roles, including Chief Strategy Officers, Chief Technology Officers, and Heads of Product Development, among others, and possessed direct knowledge of their firm’s generative AI initiatives. An independent third party administered the survey to ensure a balanced and representative sample.