Technology / Business

Add as a preferred source on Google

Financial Firms Test Multi-Agent AI, but Only 10% Have Enterprise-Wide Orchestration

Capgemini found 59% of surveyed firms were piloting or deploying the systems, while more mature firms reported higher rates of coordination and governance.

Updated 2 min read
Financial services staff review an AI workflow inside an operations floor in Paris, France. (AI-generated image / EconoTimes)
Financial services staff review an AI workflow inside an operations floor in Paris, France. (AI-generated image / EconoTimes)

Financial services firms are testing multi-agent artificial intelligence faster than they can coordinate it across their businesses, Capgemini found, highlighting a gap between trials and companywide deployment.

Capgemini’s World Cloud Report for Financial Services 2027 found that 59% of surveyed firms were piloting or deploying multi-agent systems, while 10% had reached enterprise-wide orchestration that supports autonomous business execution. The report draws on the 2026 Global Financial Services Executive Survey and interviews with senior executives from financial institutions and hyperscalers. The survey included 1,100 financial services leaders across 14 markets from June to July 2026.

Four in five executives ranked technology debt reduction among their priorities for preparing for agentic AI. Capgemini grouped two investment concerns in one estimate: roughly half of firms could not consistently show returns and lacked clear criteria for deciding whether to keep funding initiatives. To manage token costs and performance, 90% were using or planning to adopt intelligent model routing, and most expected to do so within a year.

Cloud governance is another focus. Hyperscalers were the preferred partners for governing multi-agent deployments among 52% of banks and insurers, Capgemini said. Regulatory requirements, data residency and compliance credentials influenced cloud selection for 50% of executives. Security and identity management were cited by 47%, while 39% pointed to auditability and traceability.

About 5% of firms qualified as agentic AI maturity leaders in the report. Within that group, 51% had enterprise-wide orchestration, against 8% of other firms. Automated enforcement or policy-as-code backed governance at 94% of leaders, compared with 43% elsewhere. Eighty-three percent of leaders had established or improved accountability for outcomes, versus 25% of other firms.

Capgemini projects that by 2028 the number of executives using business outcomes as their main AI funding criterion will rise 141%, while the number emphasizing efficiency will fall 42%. The company also said three-quarters of respondents expect agents to become important to customer engagement within three years, including account servicing, loan origination, underwriting and product recommendations.

Related articles

View all