Who Owns the Customer Conversation in the Age of AI? 

The next banking shift is already underway

In my previous blog on conversational banking, I explored how its implementation is reshaping bank customer interactions. What began as an evolution in user experience is now becoming something far more fundamental.

Customers are changing how they engage with digital systems. They are asking questions, expressing intent, and increasingly expecting systems to respond with clarity and relevance. This shift is driven in large part by the rapid adoption of large language models. These tools have made conversational interaction intuitive, widely accessible, and often preferred.

In parallel, Open Banking and Open Finance have made financial data more portable and interconnected. Customers can access multiple providers, compare products, and move their data across platforms with relative ease. This openness has expanded choice, but it has also introduced complexity and paved the way for AI agents to own an ever-growing share of the financial advice interactions. The challenge is no longer access to data. The challenge is making that data meaningful in real time, to stand a better chance to own the interaction.

As conversational interfaces become more common, they begin to shape expectations in financial services. Customers want to ask their bank a question and receive a crisp response that reflects their situation, their goals, and their constraints. This requires systems that understand and potentially act.

The shift toward conversational banking is therefore not just about interaction. It is about control, context, decision-making and action.

The strategic decision banks must make

Banks are now approaching an inflection point.

One path is to build and maintain their own conversational experiences, and embed them across their ecosystem, digital channels and product experiences (from onboarding to servicing to closure). This approach allows banks to retain the interaction and not hand it over to consumer LLMs, keep the data flywheel going (data from interactions > better understanding of needs and preferences > better product and experiences > better retention and acquisition > more data.

The alternative is to ensure that their products and services are accessible within third-party AI ecosystems. Customers are already spending time in environments powered by ChatGPT, Claude, Gemini, and others. These platforms are becoming the starting point for discovery, comparison, and increasingly, decision-making.

But each path has unique implications.

Building in-house provides control and alignment with existing systems, but it may limit reach. Integrating into third-party ecosystems increases visibility and accessibility, but it introduces new dependencies and reduces direct ownership of the interaction.

At the center of this decision is the question of intermediation. If the customer journey begins outside the bank, then the bank is no longer the primary interface. Over time, this can shift where value is created and where relationships are maintained.

The decision is not binary, but it is strategic. Banks will need to determine how they participate in an environment where the interface is becoming more distributed.

When AI begins to guide financial decisions

As conversational systems mature, their role is expanding beyond answering questions. They are beginning to guide decisions.

An AI assistant can already analyze a user’s financial behavior, identify patterns, and surface relevant insights. The next step is to move from insight to recommendation, and from recommendation to action.

In this emerging model, an AI system could evaluate multiple financial products across institutions. It could compare rates, terms, and eligibility requirements. It could then present a recommendation based on the customer’s goals and preferences.

In such a scenario, banks interact less with customers directly and more with the systems that represent them. These systems are designed to optimize outcomes for the user. They operate continuously and respond quickly to new information.

This introduces a new dynamic. Banks may need to respond in near real time, adjusting offers, pricing, or conditions as part of a competitive environment where decisions are mediated by AI. The interaction becomes less about presenting a static product and more about participating in a dynamic process.

This does not eliminate the role of the bank, but it changes how that role is expressed.

What AI still depends on

Despite the rapid progress of AI, there are foundational capabilities that remain outside its scope.

AI systems do not own financial products. They do not hold deposits, extend credit, or manage regulatory obligations. They operate on top of information, but they rely on underlying institutions to deliver outcomes.

This distinction is important.

Banks continue to hold two essential capabilities. The first is context. Banks maintain detailed knowledge of customer relationships, financial history, and product usage. This context is built over time and is grounded in real transactions and interactions.

The second is execution. Banks have the infrastructure to create products, set pricing, enforce eligibility, and fulfill transactions in a compliant manner. These are not abstract capabilities. They are operational and regulated.

AI systems can enhance decision-making, but they require access to structured context and reliable execution to be effective in financial services.

As conversational interfaces become more prominent, the importance of these capabilities increases. The interface may change, but the underlying requirements for trust, accuracy, and delivery remain.

From insight to execution

The industry has invested heavily in generating insights. Dashboards, alerts, and recommendations have improved visibility for customers and institutions alike.

However, insight alone does not drive outcomes.

Execution is what translates understanding into value. In a conversational environment, this means being able to respond immediately to customer intent. It means moving from identifying an opportunity to acting on it within the same interaction.

This creates new demands on banking systems.

Pricing must be adaptable. Products must be configurable. Offers must be generated and updated quickly. The infrastructure must support continuous iteration.

These capabilities are becoming more relevant as decision cycles shorten. When interactions are mediated by AI, expectations for responsiveness increase. Systems need to keep pace.

This is where many banks encounter friction. Legacy processes, manual steps, and fragmented systems make it difficult to respond dynamically. Bridging that gap is a key part of preparing for a conversational future.

The role of context protocols and connected intelligence

To participate effectively in AI-driven interactions, banks need a way to connect their internal capabilities with external conversational systems.

Model Context Protocol, or MCP, represents one approach to this challenge. It provides a structured way to link conversational interfaces with the data, rules, and decisioning systems that exist within the bank.

Through MCP, a bank can expose specific capabilities in a controlled and contextual manner. For example, an AI system could access enriched transaction data to better understand customer behavior. It could reference product rules to determine eligibility. It could retrieve pricing information and evaluate potential actions.

This approach allows banks to maintain control over how their capabilities are used, while enabling AI systems to operate with greater accuracy and relevance.

It also supports flexibility. Banks can work with different AI providers, integrate new technologies over time, and evolve their architecture without being tied to a single interface.

At Zafin, we see this as part of a broader shift toward connected intelligence. The goal is to ensure that wherever a conversation takes place, it is supported by reliable data and actionable capabilities.

Reclaiming the relationship through intelligence

As interfaces evolve, the nature of the customer relationship evolves with them.

A conversational environment creates opportunities to reintroduce understanding into financial interactions. When systems can interpret intent and respond with context, the experience becomes more aligned with customer needs.

This opens the door for banks to play a more active role in supporting financial wellbeing. They can anticipate needs, provide timely guidance, and enable decisions that reflect individual circumstances.

Trust becomes central in this model. Customers need confidence that the guidance they receive is accurate and aligned with their interests. This requires transparency, consistency, and strong underlying systems.

Banks that invest in context and execution are better positioned to deliver on this promise. They can participate in conversations with clarity and follow through with action.

A future shaped by conversational distribution

The way financial products are discovered and accessed is changing.

Conversations are becoming a primary entry point for engagement. Whether through bank-owned channels or third-party environments, the interaction begins with a question and moves quickly toward a decision.

This has implications for how products are positioned, how services are delivered, and how relationships are maintained.

Banks that prepare for this shift will focus on strengthening their underlying capabilities. They will ensure that their data is structured, their systems are connected, and their processes support rapid response.

At Zafin, our perspective is that this foundation matters more than the interface itself. By enabling banks to manage pricing, products, and offers dynamically, we help them respond to changing expectations with confidence.

The future of banking distribution is conversational, and in that future, conversations may take place in many environments. What remains constant is the need for accurate context and reliable execution.

Banks that can provide both will continue to play a central role in their customers’ financial lives.