Banks' focus on cost savings obscures agentic AI's bigger payoff

The data: Financial institutions are adopting agentic AI quickly, but demonstrating a return on their investments has proven challenging, according to Capgemini's “World Cloud Report for Financial Services 2027.”

  • 59% are piloting or deploying multi-agent systems, but only 10% have connected AI agents across their organizations to support autonomous business functions.
  • 48% of firms say they cannot consistently demonstrate ROI from agentic AI initiatives.
  • 51% lack established processes for evaluating whether AI projects deliver enough value to justify further investment.

Customer service tops the list of banking workflows targeted for multi-agent execution over the next 18 months (55%), followed by KYC and compliance operations (49%) and fraud detection (43%), per Capgemini.

But many financial institutions are still operating disconnected AI pilots, limiting their ability to coordinate agents across business functions and demonstrate enterprise-wide value, per Capgemini.

And while the cost of running AI models has declined, more complex agentic workflows require more AI processing—offsetting some of those savings.

Implications for banks: Banks are prioritizing applications with easily quantifiable returns, such as reducing call center costs or automating back-office processes.

But as bank agentic use of AI evolves, they'll need to demonstrate how the technology delivers broader business value, including better customer experiences and stronger relationships.

For example, AI agents that proactively identify financial needs, help customers avoid fees, or provide personalized guidance could strengthen customer loyalty and ultimately support deposit growth and cross-selling opportunities. But those benefits may take longer to translate into measurable financial returns than operational cost savings.

That doesn't mean abandoning ROI discipline, though. Instead, banks should establish different measures of success for different applications. While operational AI can be evaluated through cost reductions and productivity improvements, relationship-focused initiatives should also be measured through customer satisfaction, engagement, and problem resolution—with longer-term assessments of retention, deposit growth, and product adoption.

This content is part of EMARKETER’s subscription Briefings, where we pair daily updates with data and analysis from forecasts and research reports. Our Briefings prepare you to start your day informed, to provide critical insights in an important meeting, and to understand the context of what’s happening in your industry. Non-clients can click here to get a demo of our full platform and coverage.

You've read 0 of 2 free articles this month.

Get more articles - create your free account today!