Industry

Itaú Launches ia.i, a Conversational AI Layer for 70 Million Brazilian Customers

Brazil's largest bank deploys OpenAI models across checking, credit, cards, and insurance in a unified voice and text interface, prioritizing simplicity over technical showmanship.

Last verified:

Itaú’s Unified Banking AI Strategy

Itaú Unibanco, Brazil’s largest financial institution, launched ia.i, a conversational AI assistant accessible through text and voice interfaces, consolidating customer interactions across checking accounts, credit cards, insurance, and other services into a single environment. According to the OpenAI Blog, the deployment reaches approximately 70 million customers, with an initial rollout phase encompassing 300,000 users for iterative learning and refinement.

The assistant was built on a multi-model approach using OpenAI’s GPT lineup rather than a single foundational model. Itaú CTO Carlos Eduardo Mazzei explained that this architecture allows the bank to optimize for different use cases: cost-efficient, low-latency models handle routine account inquiries, while more complex reasoning models support investment guidance and financial decision-making.

Why Simplicity Over Technology Showcase Matters

Itaú’s design philosophy prioritizes what Mazzei calls “invisible technology”—solving tangible financial problems without making the AI itself the centerpiece of the customer experience. According to the OpenAI Blog, research conducted by Itaú and Grupo Consumoteca found that 80% of Brazilians want better organization of their financial lives and 65% seek AI-powered financial guidance systems. The bank’s strategy responds directly to this demand by positioning ia.i as a financial copilot offering ongoing guidance rather than a novelty interface.

The solution rests on three operational pillars: interactivity (seamless conversation), multimodality (text, voice, image, and clicks in one experience), and individualization through Customer DNA—a proprietary data layer that maps customer needs, spending habits, and life stages while applying governance and responsible AI safeguards.

From Pilot to Scale

Itaú introduced Inteligência Itaú capabilities starting in 2024 for specific financial use cases. ia.i represents the next evolution: a comprehensive conversational layer spanning the bank’s entire product suite. The initial 300,000-customer cohort functions as a learning population, allowing the bank to measure real-world adoption patterns and direct financial benefit per user before broader expansion.

Why This Matters

Itaú’s deployment signals a shift in enterprise AI strategy: rather than racing to integrate cutting-edge models into flashy interfaces, leading financial institutions are prioritizing embedded, contextual AI that compounds personalization over time. The Customer DNA layer—enriched progressively by GPT models—suggests a model for scaled financial services personalization where AI quality is measured by reduction in customer decision friction and support volume, not benchmark scores or feature visibility.

For OpenAI, the deployment validates multi-model architecture in regulated financial services, where cost-per-transaction and reasoning depth both matter. For competing model providers, it underscores that enterprise wins in banking depend on operational fit (governance, latency, cost optimization) as much as raw capability.

Frequently Asked Questions

What is ia.i and who can use it?

ia.i is Itaú's conversational AI layer accessible via text or voice, serving approximately 70 million customers. It consolidates interactions across checking accounts, credit, cards, insurance, and other services into one environment, with escalation to human support when needed.

Why did Itaú choose OpenAI models over a single model?

According to Itaú CTO Carlos Eduardo Mazzei, OpenAI's model range enabled the bank to optimize across different use cases—from cost-efficient, low-latency solutions for simple queries to complex reasoning models for investment advice and financial grounding.

How does Itaú personalize responses at scale?

The bank built 'Customer DNA,' a financial-context layer that maintains information on customer needs, habits, and life stages. GPT models help maintain and enrich this representation, making interactions progressively more relevant under governance and responsible AI controls.

#generative-ai #banking #enterprise-deployment #brazil #openai