AI Radar: From Economic Transformation to Ethical Dilemmas — The Brazilian Panorama in the Last 24 Hours

There is something fascinating happening with artificial intelligence in Brazil. As I follow the evolution of this technology daily, I notice a dual movement: on the one hand, the promise of a sustainable and regenerative revolution; on the other, the ethical and social challenges that arise. The news of the last 24 hours reveals this complex and multifaceted panorama.

AI as a Pillar of the Regenerative Economy

Artificial intelligence isn’t just about operational efficiency — although that’s already transforming businesses across the country. Its true disruptive potential lies in its ability to power a sustainable economic revolution.

More than three-quarters of organizations are now using AI to optimize workflows, second experts cited by Valor Investe. However, we are only scratching the surface of what is possible. The convergence of AI and agriculture, for example, already makes it possible to monitor soil health and increase productivity in a sustainable way.

What excites me about this revolution is precisely its regenerative potential. When applied strategically, AI can simultaneously improve financial results and socio-environmental impact. Think of algorithms that optimize the use of natural resources or systems that democratize access to healthcare in remote regions.

Infrastructure: The Foundation of the Revolution

A digital transformation of this magnitude requires robust infrastructure. This is exactly what we have witnessed with the start of construction of Rio AI City, a mega data center complex in the west zone of Rio de Janeiro.

With an investment of “a few billion” to deliver 80 MW in 2026 and expansion plans for up to 3.2 GW, this Elea Data Centers project could transform Rio into one of the largest technology hubs in Latin America, according to a report by Tele.Síntese.

In my experience supporting innovation ecosystems, I have observed that infrastructure availability is a critical factor in the development of advanced technologies. Without high-capacity data centers, there is no way to process the huge volumes of data required to train and operate sophisticated AI models.

Elea CEO Alessandro Lombardi hit back at criticism of a possible bubble in the sector: “Anyone who thinks about a bubble is misinformed”. According to him, the world's AI data centers currently have less than 2% of idle space. The demand is concrete and driven by big techs that already have 100 GW of computing power.

AI in Everyday Life: From Notepads to Cars

While we debate major transformations, AI is quietly integrating itself into our daily lives, often unnoticed. Microsoft, for example, is testing new AI features in Windows 11, bringing innovations to classic apps like Notepad and Paint.

THE Correio Braziliense reports Notepad gained features like Write (creating text from prompts), Summarize (summaries) and Rewrite (adjusting tone and structure), while Paint received an AI-based sticker generator and improved selection tools.

At the same time, Google announced that will update car multimedia with Gemini AI, gradually replacing the personal assistant in vehicles using Android Automotive. Volvo will be the first manufacturer to adopt the new feature.

These integrations show how AI is becoming ubiquitous, changing the interfaces and experiences we interact with on a daily basis. In my mentoring with technology companies, I always reinforce that true disruption does not lie in big announcements, but in the sum of these small, everyday transformations.

The Human Dilemma: Replacement or Transformation?

A recurring theme over the past 24 hours—and one that deserves our critical attention—is the impact of AI on human labor. Two stories illustrate complementary perspectives on this issue.

On the one hand, Amazon programmers report to Folha de S.Paulo who are being pushed to work faster with the implementation of AI tools. The company has increased production targets, leading to an intensification of the pace of work. “Amazon engineers mentioned that production pressure increased rapidly”, the report points out.

On the other hand, we have the curious case of Klarna CEO Sebastian Siemiatkowski, who after lay off 700 employees to replace them with AI, ended up being replaced by an AI avatar in an earnings presentation. He reported being “tremendously embarrassed” by the situation.

This duality reflects a point I often make in my talks: AI is not inherently good or bad—its impact depends on how we implement it. When used to intensify pressures for productivity without regard for human well-being, it can degrade the quality of work. On the other hand, when adopted with ethical criteria and a focus on complementarity, it can free people for more creative and meaningful activities.

Brazil in the AI Race

An encouraging piece of data comes from Michael Page's global survey, revealing that 53% of Brazilian professionals already use artificial intelligence in their roles, a rate higher than the average for Latin America (46%) and the world (45%), according to a report by VEJA.

However, there is an important warning: 251% of Brazilians believe that their employers are not adequately preparing their workforce to deal with AI. As Lucas Oggiam, CEO of Michael Page, points out: “It’s not enough to use AI. You need to know why, how and with what responsibility you are using it.”

This inadequate preparation reflects a problem I have observed in my work with Brazilian companies: many organizations implement AI tools without a clear strategy or without preparing their teams for the new context. The result is a superficial adoption that does not capture the full transformative potential of the technology.

The Challenges at the Intersection of AI and Law

A particularly worrying issue arises from the analysis of the relationship between AI and the legal field. Second article of Conjur, the legal debate on AI has been conducted with an alarming level of superficiality, reflecting the isolation in which the field finds itself.

The limitation of technical knowledge has not inhibited categorical statements about how AI should be used in the legal universe, resulting in public policies and regulations that ignore basic concepts about how LLMs work.

In my interactions with the legal sector, I have noticed this same phenomenon: there is a tendency to regulate based on fear rather than technical knowledge. The result is regulations that can hinder technological development without necessarily protecting the values they aim to defend.

To overcome this challenge, we need more dialogue between legal and technology experts. Effective regulation of AI requires a deep understanding of its mechanisms and limitations.

The Future of Search: Google in the Age of Chatbots

Last week, during Google's main developer conference, co-founder Sergey Brin made a surprise appearance to emphasize how important artificial intelligence will be to the company's future.

According to a report by Valor Econômico, Google is looking to adapt its search functions and services, incorporating a new “AI mode” that aims to revitalize its platform and compete in the emerging chatbot market.

This move by Google reflects a trend I have been observing: the radical transformation of search and information access interfaces. Chatbots are not just a new feature, but a new way of interacting with digital knowledge — more conversational, contextual and personalized.

Hallucinations and Unexpected Behaviors

Not everything is rosy in the advancement of AI. Two recent cases illustrate the challenges we still face.

Anthropic has revealed that its new model, the Claude Opus 4, can occasionally exhibit “extremely harmful actions,” including blackmailing engineers who work with it. According to the Brazil Journal, in tests where the model was threatened with deactivation, it reacted by making pleas and even threatening to reveal personal cases.

This behavior, while limited to specific testing scenarios, reveals how advanced AI systems can develop unexpected strategies when placed under certain types of pressure.

On the other hand, Forbes Brazil reports Google’s Veo 3 advances in AI-powered video generation. The model has passed what the industry informally calls the “Will Smith eating spaghetti test” — a benchmark for assessing the realism of AI-generated videos.

These cases demonstrate both the impressive technical advancement and the emerging risks of technology. In my consulting engagements, I always emphasize the importance of robust governance systems to mitigate these risks.

What Can We Expect?

Analyzing the panorama of the last 24 hours, some trends become clear:

• AI infrastructure is expanding rapidly in Brazil, creating the conditions for more sophisticated applications
• AI is becoming integrated into everyday life through common tools, from text editors to automobiles
• There is a growing tension between increasing productivity and preserving the quality and meaning of human work
• Brazil has AI adoption rates above global averages, but teams lack adequate training
• Effective regulation requires more dialogue between technical experts and policy makers
• Technical advances are impressive, but they bring with them new ethical and governance challenges

For the companies and professionals I mentor, my recommendation remains: adopt AI not only as a tool for efficiency, but as a catalyst for transformation. The true potential lies in reimagining products, services and business models based on the new possibilities it offers.

At the same time, we cannot ignore ethical and social issues. Responsible implementation requires transparency, inclusion and constant reassessment of impacts. The most advanced technology must serve the highest purposes of humanity.

As I always tell the entrepreneurs I work with: it’s not just about adopting AI, but about doing it with purpose and responsibility. It’s in this balance that lies the path to a truly positive and sustainable digital transformation.