Itaú Builds Generative AI on OpenAI Models: ia.i, IARA, and a Plan to Reach 70 Million Customers

Itaú is building generative AI for its entire customer base by combining a range of OpenAI models with its own governance layer, rather than relying on a single model. The Brazilian bank's conversational product, ia.i, reached approximately 300,000 customers in its first phase and is expected to be available to the full customer base by the end of 2026.

What Itaú announced

Itaú introduced ia.i as a complete conversational layer for banking, activated by text or voice, and described the architecture behind it: a Customer DNA context layer, a governance and orchestration platform called IARA, and an internal "AI First" strategy that extends the same intelligence layer to the bank's own specialists and service teams.

The bank frames the initiative around simplicity rather than model sophistication. Carlos Eduardo Mazzei, CTO of Itaú, describes the goal as making the technology almost invisible to the customer.

"Much of it is about making something simple, direct, and useful, delivering a direct benefit to the customer, and making the technology almost unnoticeable." —Carlos Eduardo Mazzei, CTO, Itaú

The design principle: make the technology disappear

Itaú's stated standard for the product is that a solution must solve a real everyday problem and be direct, simple, and memorable. In this framing, technology is the means and never the centerpiece of the customer experience.

The bank cites demand for this kind of support. A study by Itaú in partnership with Grupo Consumoteca found that 80% of Brazilians want to better organize their financial lives and seek more effective support for decisions, and that 65% would like technologies that act as a financial copilot, providing ongoing guidance.

ia.i: one conversational layer across banking products

The journey began in 2024, when Itaú introduced the first Inteligência Itaú capabilities for specific financial needs. ia.i is the next stage: a single environment that brings interactions related to checking accounts, credit, cards, insurance, and other services together, and connects customers to specialized human support whenever needed.

The breadth of those use cases drove the decision to use multiple models rather than one.

"OpenAI's models were very important in delivering this experience, because they were able to operate across this range of solutions. From the simplest ones, which need to be efficient in cost and latency, to more complex solutions, where reasoning and all the necessary grounding are required to provide an appropriate answer." —Carlos Eduardo Mazzei, CTO, Itaú

According to Itaú, having a range of models — instead of a single model — is what made it possible to deliver the complete solution in one product. The experience rests on three principles: interactivity, multimodality (text, voice, image, and clicks in the same experience), and individualization.

The first phase covered approximately 300,000 customers, which the bank treated as a learning period about the real relationship between customer and product.

"We can understand how customers begin to use this product more deeply and see how it can directly benefit each person's daily financial life." —Carlos Eduardo Mazzei, CTO, Itaú

ia.i spans the customer journey, from a question about an expense in an account statement to investment advice and convenient transactions on WhatsApp.

"The financial world often has products that are not so straightforward and need this context, this simpler relationship with money." —Carlos Eduardo Mazzei, CTO, Itaú

Customer DNA: financial context for approximately 70 million customers

At the center of the ia.i experience is Customer DNA, a layer of financial context that helps Itaú understand the needs, habits, and life stages of approximately 70 million customers. Under governance and responsible AI controls, GPT models are among the components that help maintain and enrich this representation, making interactions progressively more relevant. Itaú states that GPT‑5.x models are part of Customer DNA.

IARA: the governance layer that makes scale possible

IARA is Itaú's governance and orchestration platform. It gives the bank's teams controlled access to different models in line with internal policies, and OpenAI is among the model providers available in that environment.

IARA addresses three problems at once: cost control per token, accuracy of responses, and application of responsible AI policies. Itaú states that this combination is what allows the bank to move quickly while maintaining control in a regulated sector.

"IARA is a fundamental platform, because it ensures we can have governance and observability in a highly regulated environment." —Carlos Eduardo Mazzei, CTO, Itaú

"It is essential that this information be preserved, remain within the bank's context, and access models in a secure, governed way, with an understanding of token-use efficiency so that the entire solution is economically viable for the bank." —Carlos Eduardo Mazzei, CTO, Itaú

Responsible AI as a product requirement

Itaú describes Responsible AI as a discipline the bank treats with great care, and says responses to customers must align with the institution's values and culture. That includes reducing any risk of bias that may arise in frontier models. Responsible AI principles guide the entire lifecycle of solutions, from development to continuous monitoring, aiming to ensure safety, transparency, ethics, and regulatory compliance as the technology scales.

"When we talk about Responsible AI at the bank, it is a discipline we take great care with. We need customers to interact with the bank and our responses to align with our values and culture." —Carlos Eduardo Mazzei, CTO, Itaú

The legal department: 90 million documents, 2 million tasks per month

IARA supports internal solutions as well as customer-facing ones. In Itaú's legal department, the platform automates document reading, analysis, and interpretation. According to the bank, the solution operates on a repository of more than 90 million documents and executes more than 2 million tasks per month, with efficiency gains of 20% to 80% in stages of the department's processes.

AI First on the inside, too

Itaú's AI First strategy extends beyond end customers. The same intelligence layer that supports ia.i also assists specialists, managers, and customer service teams.

The bank describes a range of internal uses, from automating simple tasks to performing work that was previously unfeasible, while adding value to customer relationships: summaries before meetings, meeting preparation, understanding a specific issue, and developing plans and alternatives that were previously less obvious and time-consuming.

Itaú states significant expectations for this internal implementation: three times more time available for customer relationships, service up to twice as fast, and conversations up to three times more effective.

"This set of tools in the hands of relationship managers allows us to focus on what matters most: creating value for our customers." —Carlos Eduardo Mazzei, CTO, Itaú

Results overview

ia.i reached approximately 300,000 customers in its first phase, covering everything from convenient transactions to investment advice, activated by text or voice.
The IARA platform makes different models available to the bank's teams with governance, observability, cost control per token, and enforcement of Responsible AI.
The legal solution operates on more than 90 million documents and executes more than 2 million tasks per month, with efficiency gains of 20% to 80%.
Using a range of models made it possible to serve both cost- and latency-sensitive cases and those requiring deeper reasoning; GPT‑5.x models are part of Customer DNA, which provides financial context for approximately 70 million customers.

Lessons for leaders

Itaú distills its experience into five points for other organizations building generative AI at scale:

Measure the experience by its simplicity, rather than sophistication. The goal is for the technology to go unnoticed and the benefit to be apparent.
Choose a range of models, rather than a single one. Real journeys combine low-latency interactions with decisions that require reasoning.
Build the governance layer before scaling. Observability, cost control, and responsible AI make scale viable in regulated sectors.
Treat bias and cultural alignment as product requirements. Responses to customers must reflect the institution's values.
Apply the strategy internally, too. Giving time back to teams frees up human relationships where they matter most.

Tips from the implementation

Start with concrete everyday problems, such as a question about an account statement or a simple transaction.
Launch to an initial customer group and use it to learn actual behavior before expanding.
Centralize model access on one platform to maintain control over costs and quality.
Monitor token efficiency from day one — economic viability is part of the design.
Combine conversation and navigation: not every action is best handled through natural language.

Next steps

ia.i continues to evolve. Itaú expects to make the experience available to its entire customer base by the end of 2026. Planned capabilities for the next stages include product simulations and sign-ups, as well as recurring financial analyses of income, expenses, and habits.

Mazzei places the initiative in the context of the bank's longer history of navigating transformation, with customers' habits and expectations constantly changing.

"Leading the AI era is about responding faster to our customers' needs. It is about understanding how technology can add value and help customers engage with financial products and services in a much simpler, more direct way that helps them achieve what they are seeking." —Carlos Eduardo Mazzei, CTO, Itaú