AI, responsibility, and the model we are building

Artificial intelligence is moving incredibly fast and I think that, as a society and as companies, we need to talk about it with more calm, more responsibility, and less euphoria.

This is not about being against technology. Quite the opposite. AI can help us enormously: automate repetitive tasks, improve processes, reduce errors, make better decisions, and free up time for higher-value work. But we also have to accept that not every technological advance is neutral.

We often talk about AI as if it were just another tool. And while that is partly true, I do not think we fully understand yet the real shift it is bringing and the one that is still ahead. It is not only an incremental improvement. It is a change of model. And when the model changes, the rules should change too.

When a company starts replacing human work with intelligent agents or robots in production, it is not only optimizing costs. It is redefining how value is created. Productivity may increase, yes, but people can also be left behind if this is not managed responsibly.

That is why we will need to talk seriously about regulation. Not as a brake, but as a balancing mechanism. If a growing share of value is generated without direct human intervention, we need to think about new forms of contribution: taxation linked to automation, redistribution mechanisms, or even basic income.

But responsibility should not begin with the law. It should begin inside companies themselves.

Any organization adopting AI should have a clear set of principles, almost a manifesto: understand what it automates and why, what impact it has on people, how it protects data, how it avoids bias, and which decisions should never be fully delegated to a machine. Not everything that can be done should be done.

And there is another point that is often overlooked: environmental impact.

AI is not intangible. Behind it there are data centers, energy consumption, hardware, cooling... Every model we train and every system we scale carries a real cost for the planet. That creates responsibility there as well: looking for more efficient solutions, optimizing what we build, and avoiding the habit of using AI just because we can without thinking about the impact.

Maybe today we can still trust that regulation will eventually come or that the market will correct itself. But we are probably still underestimating the real scale of what is coming. And precisely because of that, responsibility cannot be optional or something we postpone for later.

It has to be part of how we design, how we decide, and how we build from now on.

In the end, the challenge is not only to move faster or become more efficient. It is to do it well. To build a model where technology creates value without leaving anyone behind and without pushing the cost onto the environment.