It has been just over three years since Generative Artificial Intelligence became mainstream with the public launch of ChatGPT. This is a good time to look back and understand what has worked, what hasn’t, and what companies can do to leverage this technology effectively and securely.

Access the whitepaper

Generative AI Business Applications

Let’s break it down.

What Worked?

AI finally became accessible to anyone with an internet connection. Today, more than 800 million people interact with ChatGPT every week. It also became one of the most downloaded apps of 2025, with hundreds of millions of cumulative installations, according to AppMagic.

Personal adoption of AI has been extraordinary. Users rely on it as a personal assistant, creative partner, tutor, emotional support tool, or to solve everyday questions. However, its impact in the corporate world has been different.

What Didn’t Work?

At the enterprise level, moving from pilot projects to tangible results remains the main bottleneck.

A study by the NBER (February 2026), based on a survey of nearly six thousand CFOs, CEOs, and executives from Germany, Australia, the United States, and the United Kingdom, found that 70% of companies already use AI in some way. However, more than 80% still report no measurable impact on productivity or employment. In addition, although more than two-thirds of top executives say they use AI, their average usage is only 1.5 hours (90 minutes) per week.

In other words, Generative AI has been a global success in terms of adoption, but it has not yet translated into consistent business results.

So, what can be done to change that?

Below are the three most relevant use cases of Generative AI applied in business environments, according to Nicolás Jodal, CEO and co-founder of GeneXus. Watch the video presentation here.

Generative AI Use Cases in Companies

Use Case #1: AI on Every Computer

When Bill Gates and Paul Allen founded Microsoft in 1975, they articulated an ambitious vision: “a computer on every desk and in every home.” It wasn’t just a slogan – it was a strategic north star. Although Microsoft built software, not PCs, they bet on creating platforms used by multiple manufacturers. Decades later, that vision transformed the world.

Just as computing became democratized and reached every desk, AI must also be democratized within companies. Jodal’s premise is simple: every corporate computer should have AI.

The first step is a chat. But not the public version of ChatGPT.

Organizations need a governed corporate chat that protects information and complies with internal policies. This implies role-based access control, traceability, encryption, data isolation, and secure connections to internal documents and systems. Only then can AI deliver value at scale without compromising security or compliance.

Here, a key concept emerges: “super search.” Jodal emphasizes that for the first time in history, staff can securely and privately query all the textual information the organization already owns.

“For years, companies tried to build an internal ‘Google’ for intranets and repositories. It didn’t work. Today, there is a pattern that does work: RAG (Retrieval-Augmented Generation). This approach makes it possible to query proprietary documents while respecting permissions and traceability. It’s about giving each person intelligent access to the information relevant to them and enabling new ways of working.”

Real Example

Nicolás Jodal uses specialized assistants that monitor competitors, detect changes in business models, product launches, and relevant updates. He interacts with them through specific questions and tasks. In practice, he uses several assistants, each focused on a different area.

Use Case #2: AI in Every System

Generative AI is also transforming how we interact with systems.

“Today, we use the point-and-click pattern. That model won’t disappear, but another is emerging: expressing intent. Instead of navigating menus, the user can state what they want to accomplish. The system interprets that intent and guides them to the right place. This is called intent-based navigation,” explains Jodal.

This has another important implication: many systems will need to evolve to incorporate this new form of interaction. In many cases, it is not necessary to completely rewrite systems. It is enough to add an AI layer that captures intent, guides the user, and facilitates action execution while maintaining security and data control.

Real Example

“The other day we were talking to a company that uses an ERP system with 3,240 screens. I asked them how users know which of those 3,240 screens they need to go to. No one logs into an ERP system ‘to explore.’ They log in with a specific goal: ‘I want to know how much this customer owes me.’ Today, that navigation depends on experts who know the paths or on users who browse menus until they find what they’re looking for. That’s something that can be changed by adding an AI layer. This way, the ERP can incorporate an intelligent search box: the user writes their intent, and the system takes them directly to the correct screen. This shows that Generative AI doesn’t replace the ERP. It makes it more usable,” Jodal explains.

Use Case #3: AI in Every Business Process Through Agents

This is the most sophisticated use case and the one most widely discussed: incorporating AI into a company’s business processes. Here, we are talking about Agents – systems capable of understanding intent, consulting data, coordinating actions, and executing steps within a process.

“In companies that are already digitized, much of the back-office work involves coordinating tasks: calls, emails, messages, validations, and follow-ups. Automating that coordination is difficult or requires constant human intervention. Today, with Agents, it is possible to automate those interactions. The Agent understands the status of the case, consults internal systems, drafts communications, schedules actions, and logs evidence – while respecting rules and permissions.”

How to take action

AI adoption should not depend on isolated efforts or disconnected experiments that end without real impact.

The “Golden Path”- the structured and proven route to incorporating AI without friction or chaotic experimentation- is through Globant Enterprise AI, a technology-neutral platform that can integrate with existing systems, including legacy ones, to orchestrate AI across every workflow in a secure, responsible, and measurable way.

Globant Enterprise AI provides cost governance, performance control, and flexibility in language models (LLMs), ensuring scalable and compliant adoption.

Globant Enterprise AI does not replace what you already have. It enhances it. It makes it work better, faster, and with greater control.

Discover success stories from organizations in sectors such as finance, retail, and manufacturing that are already using Globant Enterprise AI to significantly improve productivity and transform processes with Generative AI.

For more information, visit the Globant Enterprise AI website or email us at hello@genexus.com. All emails are personally read and forwarded to the appropriate teams to provide a fast response tailored to each need.

You may also be interested in reading:

26 deep dives on AI applied to businesses

GeneXus Next: Native Agentic Low-Code Development for Mission-Critical Systems

Creation and In novation in the Age of AI

Low-Code + Generative AI: Challenges and Opportunities for CIOs