New research shows how AI agents are driving value for financial services

The tipping point for AI in financial services has finally arrived, and organizations continue to see their investments paying off — supercharged by the emergence of agentic AI.

In 2025, financial institutions are scaling their use of generative AI, with AI agents becoming the pivotal next step for driving growth, creating efficiencies, and improving risk management. 

Our second-annual ROI of AI in financial services report, commissioned by Google Cloud and conducted by National Research Group, confirms that gen AI initiatives are still delivering, with 77% of financial services executives  reporting that their organization is achieving positive ROI within the first year. The survey also revealed a powerful new differentiator helping drive this momentum: AI agents — specialized large language models (LLMs) that can independently plan, reason, and perform tasks. 

The findings are based on a survey of 556 global financial services leaders with generative AI deployment within their organizations, including C-suite and senior executives across IT, marketing, and innovation functions.

You can dive into the full report (and compare it to last year’s survey) to explore all the findings, but we wanted to give you a head start on understanding how financial services institutions are capturing new value and opportunities with AI right now. Below, we’ve curated some of the key highlights. 

Agentic AI unlocks new possibilities for driving value.

AI agents can independently plan, reason, and execute complex tasks. Combined with the ability to securely connect to enterprise data and other AI agents, these capabilities are enabling organizations to go beyond simple automation and embed intelligence directly into the business. Our survey shows that financial services companies are steadily integrating AI agents to help tackle core challenges. The top use cases for AI agents in financial services are in key business areas: 

• Customer service and experience (57%)

• Marketing (48%)

• Security (46%) 

• Finance and Accounting (46%) 

• Fraud management and detection (43%) 

• Risk management (42%)

These findings suggest organizations are building confidence in driving value in higher-risk areas, paving the way for even broader transformation. In addition, nearly half (49%) of respondents report their organization plans to allocate at least 50% or more of their future AI budgets towards AI agents, underlining their emergence as the new strategic differentiator for shaping an AI-first future.

Financial services are rapidly deploying AI agents. 

While AI agents are increasingly widespread across nearly every industry, financial services stands out for the breadth of its agentic AI adoption — especially given the sector’s historically pragmatic, risk-first approach to digital transformation.

According to the survey, 53% of financial services executives reported their organizations are actively using AI agents in production, with 40% saying they have already launched more than ten. The complexity of these AI agents can span a broad spectrum. On one end are single-task agents, such as gen AI-powered digital assistants. On the other end are highly sophisticated, multi-agent systems that can work together to perform complex tasks, taking actions on behalf of users under their supervision.

This rapid uptake of AI agents is accelerating the pace of innovation in financial services. A prime example of this is the emergence of agentic commerce, where agents handle autonomous shopping, creating new revenue opportunities for merchants and payments players while improving the customer experience. To responsibly enable this shift, Google Cloud, in collaboration with a large industry coalition of financial and technology companies, recently launched the Agent Payments Protocol (AP2) — an open protocol that provides the clarity, accountability, and security framework necessary for financial institutions to manage risk and confidently unlock a broad spectrum of agentic transactions.

Strong ROI makes AI a mission-critical enterprise investment. 

The foundational value of gen AI continues to deliver compounding returns, primarily driven by tangible gains in five key areas:  

1. 1. Productivity: Among those reporting improved productivity from gen AI, financial services executives cited improvements across both IT (74%) and non-IT staff and processes (62%), with gen AI solutions also helping to improve accuracy (61%) and time-to-market (51%).
      
      
   2. Customer experience: The majority of respondents (67%) report gen AI has resulted in meaningful value-add toward customer experience.
      
      
   3. Business growth: A large majority of financial services executives (63%) continued to see business growth as a result of gen AI, and within this group, most (70%) subsequently reported annual revenue increases stemming from gen AI adoption.
      
      
   4. Marketing: Some 56% of financial services executives reported that integrating gen AI solutions had a meaningful impact on their organization’s marketing outcomes.
      
      
   5. Security: Among financial executives reporting improved security posture from gen AI, the majority reported improvements in identifying threats (81%), threat intelligence and response (79%), and reduced time to resolution (66%). 

With financial institutions making clear progress on their core AI objectives, AI budgets are following suit. Our research shows gen AI investments are surging, with most financial services executives (79%) citing increased spend as technology costs fall. Notably, 61% report their organization is now increasing their investment in gen AI initiatives, up from 58% in 2024.

The findings underscore a critical reinvestment cycle in financial services: organizations are leveraging early wins to fund the next wave of transformation. This approach creates a continuous, self-reinforcing flywheel of innovation that helps position these organizations to succeed in the age of AI.  

Data privacy and security remain top of mind.

In financial services, security isn't just a feature—it's the foundation. That's why, even as the industry races to adopt AI, leaders are ensuring the fundamentals are locked down. When evaluating large language model (LLM) providers, their priorities are clear: data privacy and security (43%) top the list, well ahead of systems integration (29%) and regulatory compliance (28%).

While this emphasis on governance underscores the complexities and inherent challenges of innovating in a highly regulated environment, it has also paved the way for success. The industry’s deep commitment to rigorous testing and piloting for model safety and soundness has yielded tangible results, allowing financial services firms to scale their AI deployments while strictly adhering to key regulatory and governance standards.

Looking forward

AI agents have arrived, and they’re more than a new advancement — they’re a pivotal force redefining how financial services institutions operate and serve their customers. As the survey data reveals, opportunity is everywhere, and the real advantage will go to organizations that take decisive action now to prudently invest in and responsibly scale agentic AI. 

Download the full report, “The ROI of AI for financial services,” to explore all our latest insights and findings.