Qualcomm champions local AI processing on devices at Rio Innovation Week 2026

At Rio Innovation Week 2026, Qualcomm advocated for Edge AI—a technology that runs artificial intelligence tasks directly on smartphones and computers, without relying on remote servers. Diego Aguiar, Qualcomm’s sales and IoT lead in Brazil, argued that this approach reduces latency, lessens the need for connectivity, and keeps sensitive data on the device itself.

Diego Aguiar: “Technical performance is a means, not an end. AI only creates value when it reduces friction, simplifies tasks, or enables experiences that were previously impossible.”

Edge AI involves running AI models locally on a user’s device. Tasks like text translation, document summarization, photo editing, and interactions with virtual assistants can now be performed without sending data to the cloud.

Benefits of local processing

According to Aguiar, the advantage isn’t just about processor speed, but about functioning naturally even in low-connectivity situations. Lower latency and the ability to keep confidential data on the device are direct benefits of this approach.

Qualcomm evaluates new AI features using three criteria: whether they solve a real problem, whether they respond quickly without draining battery, and whether users can control how their data is used. The goal is to eliminate unnecessary effort from the human experience.

From local processing to autonomous agents

The next step is AI agents—systems that can plan trips, compare prices, and organize schedules without constant supervision. However, Aguiar warns that decisions involving values, responsibility, or significant consequences should remain under human oversight.

Diego Aguiar: “Decisions involving values, responsibility, or significant consequences should remain under human supervision.”

For him, the ideal assistant isn’t one that makes all decisions alone, but one that reduces operational work and knows when to ask for confirmation or return control to the user.

Privacy goes beyond where processing happens

Aguiar stresses that processing data locally doesn’t guarantee privacy on its own. Trust requires security, transparency, and control over what information is collected, why it’s collected, and how users can limit or disable it.

Diego Aguiar: “Privacy doesn’t depend only on where processing happens. Trust also requires security, transparency, and control.”

This also applies to sensors like cameras, microphones, and location data, which help devices understand context but can become surveillance tools if they collect too much data. The line between assistance and surveillance is defined by relevance, transparency, and control.

Brazil and the adoption of conversational interfaces

Aguiar notes that Brazilians are already used to exchanging audio and messages, which makes it easier to adopt AI interfaces based on conversation rather than traditional menus and apps. However, he argues that Brazil also needs to participate in developing the technology to adapt it to local needs.

For him, using AI effectively isn’t enough; it’s essential to actively contribute to its advancement, ensuring the technology meets the country’s context.

Frequently Asked Questions

• What is Edge AI?

  Edge AI is artificial intelligence processing done directly on a phone or computer, without relying solely on cloud servers, allowing tasks to be performed locally.

• What are the benefits of processing AI on the device itself?

  This approach reduces latency, works even with limited connectivity, and keeps sensitive data on the device. More complex tasks can still use the cloud.

• What tasks can use Edge AI on phones and computers?

  Translation, document summarization, photo editing, and conversations with assistants can already be processed locally, as long as they meet criteria for usefulness, efficiency, and user control.

• What decisions should not be made by AI agents alone?

  Decisions related to career, health, safety, money, and other significant consequences should remain under human supervision.

• Does processing data on the device guarantee privacy?

  No. Privacy also depends on security, transparency, and control over how data is used.

• Why can Brazil adopt AI interfaces more quickly?

  The habit of sending messages and audio creates a favorable environment for conversational interfaces, but it’s important for the country to participate in developing the technology.

Qualcomm’s advocacy for local AI processing reinforces the idea that technology is only truly useful when it seamlessly integrates into people’s daily lives, respecting their limits and needs.