Salesforce

Agentic AI in banking and insurance: how intelligent agents are transforming financial operations

Published: 28/08/2026

Guillaume Duret

Salesforce director Consulting Practice, Inetum

Banking and insurance organizations have spent years investing in technology to improve efficiency, digitize customer engagement, and streamline operations.

get in touch

Designing with AI instead of adding AI

One lesson becoming increasingly clear across the market is that AI often falls short when treated as a feature added at the end of a project. For years, many organizations have evolved through isolated automations layered onto legacy processes. While this approach has delivered efficiency gains, it has also created environments that are difficult to maintain and scale.

Agentic AI introduces a different mindset. Rather than asking where AI should be inserted into an existing process, organizations are challenged to rethink how that process would operate if it were designed from scratch today. This distinction is particularly relevant in banking and insurance, where value depends heavily on coordinating information, responding quickly to regulatory changes, and maintaining long-term customer relationships.

That is why, even though automation remains the primary entry point today, many institutions increasingly recognize it as only the first stage. The next phase will arrive when financial services are conceived around intelligent and automated capabilities from the outset. The shift is not about automating more steps. It is about creating processes designed to work with intelligent systems from day one.

New channels have been launched, processes automated, analytics capabilities expanded, and modernization programs accelerated. Yet many institutions still face the same challenge: more technology does not automatically translate into more business value. Financial organizations continue to operate across fragmented processes, data spread across multiple systems, and digital initiatives that are difficult to scale over time. At the same time, customers expect faster, more personalized, and more consistent experiences across every channel. This raises an important question: if a significant portion of operations has already been automated, what comes next? The answer is increasingly taking shape through agentic AI, a concept gaining momentum across the financial services sector.

While generative AI marked the first wave, helping organizations accelerate tasks, create content, and improve interactions, the next stage goes further. Agentic AI introduces systems that can work within real business processes, execute actions, coordinate tasks, access multiple sources of information, and collaborate with people to achieve defined objectives. According to PwC, banks are no longer using AI solely to automate tasks or reduce costs. The focus is shifting toward generating greater value from data, improving decision-making, personalizing services, and redesigning end-to-end processes. Competitive advantage will come from embedding AI into day-to-day operations, not from accumulating disconnected use cases. The real opportunity emerges when products, services, and processes are designed from the start to operate alongside intelligent capabilities.

Automation is no longer enough

The first wave of adoption is already underway. Document processing, reporting, information analysis, case classification, commercial operations, and data quality improvement rank among the most widely deployed AI use cases in financial services today. In practice, this enables activities such as document classification and information management to evolve continuously without requiring constant rule reconfiguration.

The difference from traditional automation is that agentic AI can learn, adapt, and improve more rapidly. Whereas conventional models depend on rigid rules and ongoing parameter adjustments, intelligent agents can accelerate testing, learning, and optimization cycles within boundaries defined by the organization. Even so, the goal is not to remove people from the process. The goal is to help people work more effectively.

Customer relationships will remain human

A common assumption is that advances in AI will gradually reduce the role of people. In financial services, that perspective is overly simplistic. Financial decisions still depend on context, professional judgment, trusted relationships, and compliance with regulatory frameworks that continue to evolve. As a result, the most advanced operating models are moving toward clearly human-centric approaches.

AI can accelerate analysis, automate repetitive work, and improve response quality. People remain responsible for supervising, adapting, and evolving both processes and customer engagement models. In practice, advisors, account managers, and commercial teams can spend less time on administrative activities and more time building meaningful conversations and long-term relationships. Customers may not even notice exactly where the change is taking place.

The most profound transformation will happen inside the organization

When organizations discuss AI, the conversation often focuses on chatbots, virtual assistants, or customer-facing experiences. Yet the greatest impact may occur behind the scenes. Agentic AI is beginning to reshape how information moves through organizations, how operations are coordinated, how priorities are established, and how decisions connect across teams and functions.

Customers will ultimately experience faster service, more personalization, and reduced friction. Behind those outcomes, however, lies a much deeper operational transformation

Data remains at the center of the model

One constant runs through almost every enterprise AI initiative: data quality remains the primary limiting factor. Artificial intelligence only generates value when it operates on trusted, connected, and governed information. That is why many current initiatives focus on less visible but critical capabilities such as deduplication, normalization, document extraction, enrichment, and continuous improvement of enterprise data.

This is where Salesforce's approach to financial services becomes especially relevant. The combination of Financial Services Cloud, Data Cloud, Einstein, and Agentforce addresses a growing need across banking and insurance:

Financial Services Cloud provides industry-specific context.

Data Cloud unifies information from multiple sources.

Einstein delivers analytical and generative AI capabilities.

Agentforce enables intelligent agents to execute actions.

The value does not come from each solution independently. It comes from their ability to work together around a unified customer view.

From copilots to agents

Over the past several years, copilots have demonstrated how AI can improve individual productivity.

Agentic AI takes that model a step further. Intelligent agents are beginning to assume complete operational responsibilities, initiating actions, consulting multiple information sources, triggering workflows, and coordinating processes within defined objectives. In banking and insurance, this already translates into areas such as commercial automation, assisted case resolution, document analysis, and support for internal operations.

The same model is extending into technology teams as well. Software development, testing, documentation, and delivery automation are areas where copilots and agents are already accelerating results. While copilots help people complete tasks, agents can actively participate in executing entire business processes.

Starting small to drive meaningful transformation

Another lesson emerging across the market is that successful transformations rarely begin with massive programs. They start with focused experimentation. Building rapid proofs of concept, validating business impact, and scaling progressively allows organizations to reduce risk while accelerating learning.

This approach aligns closely with what we often see in client engagements: bringing AI into real operational processes, building capabilities incrementally, and helping organizations move from pilots to scalable operating models. Agentic AI is no longer viewed simply as the next generation of automation. It is increasingly becoming a new way of operating.

The future of financial services will be more intelligent and adaptable

The recent history of financial services has demonstrated that technology alone does not create value. Value emerges when ways of working change. Agentic AI represents a different kind of opportunity because it enables organizations to redesign processes, connect operations, and become more adaptive. Not to replace people, but to free them to focus on what remains uniquely human: understanding context, building trust, and making better decisions. Organizations that successfully integrate agents, data, and processes into a unified ecosystem will be better positioned to evolve alongside both market demands and customer expectations. One of the most significant competitive advantages of the next decade may come from that ability.

The challenge is no longer simply adopting artificial intelligence. It is embedding intelligence into processes, data, and decision-making. Organizations that begin that journey gradually, learning and scaling along the way, will be in the strongest position to capture value at scale

Request a contact