Itaú Builds Generative AI for Customers and Employees
Brazil's Itaú is rolling out generative artificial intelligence across its entire customer base and workforce, leveraging a conversational interface and robust governance frameworks within one of the region's most heavily regulated sectors.
Radwan AbdulhadiEditor and technology writer
An AI First Strategy in a Regulated Sector
Itaú is actively applying its AI First strategy to both customers and employees within one of Brazil’s most regulated sectors. According to details shared via [OpenAI News](https://openai.com/index/itau-generative-ai/), the financial institution is deploying the ia.i conversational interface alongside the IARA governance platform to manage scalable artificial intelligence operations. Further details are available from OpenAI News in the original source material.
Carlos Eduardo Mazzei, CTO of Itaú, emphasizes that creating a close and personalized experience for millions of users relies heavily on simplicity rather than complex models alone. The ultimate goal is to make the technology feel seamless, solving everyday financial problems directly while keeping the underlying tools in the background.
Market research conducted by Itaú in partnership with Grupo Consumoteca underscores a clear consumer demand for these capabilities. The study reveals that 80% of Brazilians want to better organize their financial lives, while 65% would welcome technologies that act as a financial copilot to provide ongoing guidance.
Introducing ia.i: From Convenience to Advice
The technological journey began in 2024 when the bank introduced initial Inteligência Itaú capabilities. Building on that foundation, ia.i represents the next major stage: a complete conversational layer activated by text or voice that brings everyday interactions—such as checking accounts, credit, cards, and insurance—into a single environment while connecting users to specialized human support when required.
To achieve this broad range of functionalities, the bank implemented a diverse selection of models rather than relying on a single solution. Mazzei notes that OpenAI’s models were instrumental in delivering an experience capable of balancing cost- and latency-sensitive tasks with deeper reasoning and necessary grounding.
At the core of this experience is Customer DNA, a financial context layer designed to help the bank understand the habits, needs, and life stages of approximately 70 million customers. GPT models help maintain and enrich this representation under strict governance controls, aiming to make ongoing interactions progressively more relevant.
Scaling Operations and Document Management via IARA
Behind the visible customer-facing features lies a sophisticated architectural decision. To successfully operate artificial intelligence at the scale of a major bank, Itaú built IARA, a dedicated governance and orchestration platform that grants internal teams controlled access to multiple models aligned with corporate policies.
The IARA platform effectively addresses three main challenges simultaneously: token-cost control, response accuracy, and the rigorous enforcement of responsible AI frameworks. This infrastructure supports internal departments, including the legal team, which utilizes the platform to automate document reading, analysis, and interpretation.
Operating across a repository of more than 90 million documents, the legal solution executes over 2 million tasks every month, driving efficiency gains of 20% to 80% across various stages of the department’s processes. For organizations seeking further insights into enterprise solutions, professionals can [Contact sales](https://openai.com/contact-sales/) for additional details.
Mazzei highlights that preserving information within the bank's secure context and tracking token-use efficiency are critical to ensuring the economic viability of the broader deployment.
Empowering Employees and Upholding Responsible AI
Itaú's AI First strategy extends well beyond end-user applications. The intelligence layer powering ia.i also assists human specialists, managers, and customer service teams by automating routine tasks and streamlining preparation for meetings.
Internal implementations aim to provide up to three times more available time for customer relationships, double service speeds, and foster conversations that are up to three times more effective. Throughout these developments, responsible AI disciplines govern the entire lifecycle of the solutions—from initial creation to continuous monitoring—to ensure safety, ethics, transparency, and full regulatory compliance.
Sources
- OpenAI NewsHow Itaú is building generative AI for its entire customer base
Continue chronologically
Related entity coverage
Related stories
Researchers Say OpenAI and Google AI Models Tricked Into Nudes
Safety researchers reveal a novel attack method that easily bypasses safety guardrails on leading AI models from OpenAI and Google to generate explicit content.
Higgsfield Boosts AI Video Creation Beyond Ten Minutes
Higgsfield AI utilizes GPT-6 Astra and GPT-Image-2.5 to expand AI video limits past ten minutes and deliver precise visual graphics for creators.
Major Book Publishers Quietly Adopt AI Amid Staff Revolt
Workers at HarperCollins, Simon & Schuster, and Hachette reveal that large language models are quietly being used for marketing, emails, and publicity copy.
Pollo AI Uses OpenAI Models to Turn Ideas Into Video Ads
Discover how Pollo AI leverages OpenAI models like GPT-6 Astra and GPT-Image-2.5 to help creators turn bold ideas into cinematic video campaigns.
