AI isn’t coming to liberate you: it’s coming to intensify your day (turn you into your teacher)

But the reality that drives businesses in all kinds of businesses is much more prosaic, and for that reason more dangerous: AI is not replacing millions of people with revolutions until quietly reconfigure the power relationship in the workplace.

And this reconfiguration usually comes in a very specific direction: more rhythm, more control, more pressure, more addiction…and less margin for negotiation.

Part of the phenomenon is psychological and cultural. The corporate message promises to liberate: “delegate routine” focus on areas of greatest value.

However, accumulating evidence points to the opposite effect. When a worker incorporates AI skills on a day-to-day basis, the organization tends to see it as an opportunity to reduce workload. an opportunity to raise expectations.

This overclocking work pressure is of a different magnitude, quieter: anxiety

If you do what you needed to do in two hours, they won’t give you two hours: they’ll give you more work or you’ll be listed on what is considered a “normal” yield. Productivity cannot be converted into free time; if converted to a new base line.

The result is an intensification of work: more tareas, more speed, more yoga extension, a greater feeling that you are always “for less” than the system allows.

That is the duty to work on it overclocking if it wants a different size, quieter: anxiety. Not only because of the confusion, but also because of the uncertainty of how the example will be evaluated in an environment where exit can be multiplied by an algorithm.

If the work is done in general (correspondence, tickets, receipts, information, etc.), the AI ​​turns into a metric accelerator and therefore a two-wire weapon. And if the evaluation is also automated, the circuit tightens: more tools that produce, more metrics that require more and more systems to monitor.

However, there is a particularly important element that is only discussed in public and that should sound all the alarms in Spain: growth shadow AI, the use of artificial intelligence devices is not approved by the Company, outside IT, security or regulatory radar. No words about four “geeks” trying one chatbots.

AI as it evolves tends to favor those who control goals, metrics, parameters and data.

Let’s have a habit that lasts because it is useful, fast and above all because official channels usually run late, with limitations and without understanding the real needs of the work. Empleado has tarea, a period and a free fix that “brings back”. The temptation is obvious.

The problem is that this convenience comes at a huge price: sensible things that travel where they shouldn’t. Contract forms, business information, strategic information, internal information, code, customer data, conversations… all can be copied and paid for services that companies do not guarantee.

It is for this reason that there are typical cases of accidental leaks, and it is not difficult to understand why: the worker does not think about intellectual property or trade secrets; I think about the amount of work to survive. The perverse thing is that society, without providing safe alternatives, indirectly influences this behavior.

And here it shows the most interesting and disturbing tour: it’s not just about the “use” of artificial intelligence. As for getting into it.

Each time multiple employees create their own assistants: a shoes with the response style the department uses, a knowledge base with internal documentation, a system that recreates past incidents, a “copilot” stocked with reports, presentations or information. I do it because it works.

But what do I do de factoIt is about turning the knowledge that the company has accumulated through its activities into a reusable model and data. And this opens up an explosive question: where is this knowledge when it is converted to a chatbots? To the employee who built it, to the company that created it, to the customer whose data appears in the documents, or to the hardware supplier that enables the system?

The legal feedback is usually uncomfortable but clear: much of the work an employee produces during the day belongs to the company, and inside information may constitute company secrets. in Europe protection know-how and undisclosed information have a special legal framework.

However, the real problem is not what the standard says, but how it is managed in practice when thousands of people are “distilling” and known in systems that can get out of control, duplicated, shared, or simply poorly protected.

At best, the company is exposed to leakage problems. In the first place, without knowing it, we are financing the construction of equipment that is a necessary part of human labor… with the data generated by that labor.

In this context, the promise of Fr “enhanced processing” if he seems to lack something else: the worker as an involuntary trainer of his substitute. No, because the model has to be sent back to you tomorrow until the knowledge that made it valuable is turned into an automated and scalable asset.

And for a long time it becomes more intense day by day: more tension, more urgency, more anticipation, less respite.

Of course, the conclusion is not that it is necessary to ban artificial intelligence in the workplace. That’s why you have to speak honestly about her. If the discussion is limited to how many things are destroyed, we lose the core of the same thing: the change of power. Artificial intelligence as it evolves It helps you review goals, metrics, settings and data.

We ask that workers and society also benefit from serious internal regulations, real governance, safe alternatives to shadow AI and above all the negotiation: how it is automated, with what limits, with what guarantees and how product links are distributed.

Because in this department, the AI ​​”revolution” will not be a welfare revolution. It will simply be a new world tour.

***Enrique Dans is Professor of Innovation at IE University.