The ROI of AI in financial services

This document, prepared by Google Cloud in collaboration with National Research Group, presents the results of its second annual survey of senior leaders from financial institutions on the impact and return on investment of generative AI and AI agents. The research includes 556 C-level executives (CEO, CIO, CFO, CMO, CTO, CISO) from financial services companies with over 100 employees and more than $10 million in annual revenue, distributed across 24 countries.

The report is aimed at financial institution leaders, technology and strategy teams seeking to understand how AI is transforming the financial sector and generating tangible value. The main finding reveals a significant shift toward agentic AI, with 53% of executives reporting that their organizations are already leveraging AI agents in production, and 49% allocating over 50% of their future AI budget to agents.

The document explains that AI agents are specialized language models with specific roles, context, and objectives that can independently plan, reason, and perform tasks with access to data, APIs, and other agents. Implementations range from basic chatbots to sophisticated multi-agent systems. The most adopted use cases include customer service (57%), finance and accounting (48%), security operations (46%), fraud management (43%), and risk management (42%).

The report identifies five key areas where generative AI continues to deliver proven ROI: productivity (76% report improvements), customer experience (67%), business growth (63%), marketing (56%), and security (52%). It highlights that 77% of executives report current ROI, 30% achieve production deployment of use cases in 3-6 months, and 53% of those reporting revenue increases estimate gains exceeding 10%.

Among important findings, the document reveals that organizations with comprehensive C-level sponsorship have a stronger correlation with seeing ROI (82% vs 74% without sponsorship). The top three considerations when evaluating language model providers are data privacy and security (43%), integration with existing systems (29%), and regulatory compliance (28%).

The report concludes with a practical checklist for maximizing AI agent ROI, including finding executive sponsors, demonstrating value to secure budget, creating clear governance rules, focusing on quick wins, building trust from the start, giving agents adequate tools, and investing in talent and internal education. It includes testimonials from leaders at institutions such as Scotiabank, ATB Financial, Commerzbank, Deutsche Bank, and Banco BV who are already implementing these strategies.

Key points

• 53% of financial services executives already use AI agents in production
• 77% report current ROI from generative AI initiatives, unchanged from 2024
• 49% allocate over 50% of their future AI budget to AI agents
• Top use cases: customer service (57%), finance and accounting (48%), and security (46%)
• 76% report productivity improvements, the highest impact area for generative AI
• 67% report improved customer experience, though fewer improvements in satisfaction (NPS)
• 53% of those reporting revenue increases estimate gains exceeding 10%
• Organizations with C-level sponsorship have stronger ROI correlation (82% vs 74%)
• Top priority when evaluating providers: data privacy and security (43%)
• 79% increased AI spending even as technology costs decline