TL;DR
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Capgemini’s World Cloud Report for Financial Services 2027 says 59% of surveyed banks and insurers are piloting or deploying multi-agent AI, while 10% have achieved enterprise-wide orchestration. The research identifies technology debt, governance and difficulty measuring returns as barriers to wider use.
Only 10% of banks and insurers surveyed by Capgemini have achieved enterprise-wide orchestration of multi-agent AI, even though 59% are piloting or deploying the technology, according to the consultancy’s World Cloud Report for Financial Services 2027. The figures point to a gap between experimentation and coordinated use across business operations, as institutions seek to move AI agents into core banking and insurance workflows.
The report is based on a survey of 1,100 financial services executives across 14 markets. It says firms are considering agent-based systems for activities including account servicing, loan origination, underwriting and product recommendations. Three-quarters of respondents expect AI agents to become a key part of customer engagement within three years; this is a survey finding about expectations, not a measure of current customer-facing deployment.
Capgemini identifies infrastructure, governance and return measurement as obstacles to scaling. Eight in 10 executives say reducing technology debt is a priority. Nearly half of firms cannot consistently demonstrate a return on their AI initiatives, while 51% lack a clear framework for judging whether AI investments deliver enough value to justify continued spending.
The research also describes approaches firms are taking to manage deployments and costs. Ninety percent are using or planning to implement intelligent model routing, which directs tasks to models based on an appropriate balance of accuracy and cost. Fifty-two percent identify hyperscalers as their preferred partners for governing multi-agent systems, with security, regulatory certifications, identity frameworks and auditability among the considerations in provider selection.
Why Enterprise Orchestration Is Lagging
The adoption figures suggest that trying AI agents is more common than coordinating them across an institution. Enterprise-wide orchestration involves managing how multiple agents and systems work together, rather than deploying an agent for one isolated task. For banks and insurers, that distinction matters because proposed uses touch customer service, lending decisions and other workflows where controls, security and reliable handoffs can affect customers and operations.
The financial case is also unresolved for many firms. If organizations cannot consistently measure returns or judge whether results justify spending, they may struggle to sustain investment or decide where to expand. Capgemini’s findings therefore frame the next stage of adoption as a question of governance and measurable business outcomes, not simply how many pilots are underway. The survey does not establish that agent deployment has improved performance or reduced costs across the industry.
Capgemini expects funding priorities to shift toward business outcomes: the number of executives naming outcomes as their primary AI funding criterion is projected to rise 141% by 2028, while the number prioritizing efficiency is projected to fall 42%. These are forecasts in the report, not observed changes that have already occurred.
Products Worth Considering
LG 27UP850K-W 27-inch 4K UHD IPS Monitor
Dell UltraSharp 32 U3226QER 4K Thunderbolt Hub Monitor
Dell UltraSharp 27 Thunderbolt Hub Monitor (U2726DER), QHD
As an affiliate, we earn on qualifying purchases.
From Isolated Agents to Coordinated Workflows
Capgemini’s report describes a change in focus from individual AI agents assigned to separate tasks toward multi-agent systems coordinating actions across business lines. Ravi Khokhar, the company’s global head of cloud for financial services, said firms had moved beyond experimenting with isolated agents over the prior 12 months. His assessment is an attributed characterization of the trend, rather than a separate industry-wide measurement.
The report says only a small group—5% of respondents—has established capabilities to operationalize agentic AI at enterprise scale. More than half of this group orchestrates systems across operations, technology and customer experience, compared with 8% of other firms. Capgemini also reports that 83% of these leaders link AI investment directly to business results. The report presents these figures as characteristics of the identified group; they do not by themselves establish that orchestration caused stronger results.
Cloud is presented as part of the operating foundation for these systems. Capgemini expects firms to combine internal development, packaged products and external partners: hybrid sourcing is projected to grow from 36% of firms today to more than two-thirds within three years. The projection describes anticipated sourcing choices, not a completed shift.
What the Survey Cannot Establish
The report’s figures reflect responses from executives, and the supplied material does not provide the full survey methodology, respondent breakdowns by market or institution, or detailed definitions for measures such as “enterprise-wide orchestration.” It is therefore unclear how results vary by country, institution size or type of financial business.
Several figures describe expectations and plans rather than completed activity, including projections for customer engagement, funding priorities and hybrid sourcing. The research also does not quantify the financial returns of multi-agent AI or establish whether the firms with more orchestration achieve better outcomes because of it. The 10% orchestration figure is a reported survey result, not a regulator’s assessment or an independently verified count of all banks and insurers.
Scaling Beyond Pilots
Capgemini’s report points to technology debt reduction, governance, security and return measurement as priorities for institutions seeking to expand deployments. The near-term test for firms will be whether they can connect agents across workflows while maintaining oversight and showing why continued spending is warranted.
The report projects that customer-facing uses and hybrid sourcing will become more common over the next three years, but it does not specify implementation dates for individual institutions. Further evidence will be needed to show whether those plans become operational deployments, how firms assess their performance and whether wider orchestration produces measurable business results.
Key Questions
How many banks and insurers are using or testing multi-agent AI?
Capgemini says 59% of surveyed institutions are piloting or deploying multi-agent AI. The figure combines pilots and deployments; it does not mean all are operating the systems at scale.
How many have enterprise-wide orchestration?
Ten percent of respondents have achieved enterprise-wide orchestration, according to the report. Capgemini does not provide a detailed definition of that measure in the supplied material.
What is stopping institutions from scaling the technology?
The report identifies technology debt, governance and difficulty demonstrating returns as key challenges. It says eight in 10 executives prioritize reducing technology debt, nearly half cannot consistently demonstrate returns, and 51% lack a clear framework for evaluating AI investments.
Which financial services tasks could use AI agents?
The report names account servicing, loan origination, underwriting and product recommendations as areas expected to see greater adoption. Three-quarters of respondents expect agents to become a key part of customer engagement within three years.
Does the report show that multi-agent AI is delivering financial returns?
No. Capgemini reports that many firms have difficulty demonstrating returns, but the supplied findings do not quantify industry-wide financial gains or establish that orchestration causes better performance.
Source: finextra.com
