AI has no idea why someone might be struggling
En los despachos de miles de empresas estadounidenses, una silenciosa transferencia de autoridad está en marcha: los gestores humanos ceden a las máquinas el peso de decisiones que moldean vidas. Una encuesta de Resume Builder revela que dos tercios de los directivos consultan inteligencia artificial para decidir quién pierde su empleo, situando a cientos de miles de trabajadores ante el veredicto de algoritmos que no conocen el contexto, la trayectoria ni la humanidad de cada persona. No es el apocalipsis laboral que algunos vaticinaron, sino algo más sutil y quizás más inquietante: la erosión gradual del juicio humano en el corazón mismo de las relaciones laborales.
- El 66% de los directivos encuestados ya usa IA para decidir despidos, y un 2% delega la decisión final al algoritmo sin ninguna revisión humana.
- Herramientas de uso cotidiano como ChatGPT, Copilot y Gemini se han convertido en árbitros silenciosos de ascensos, aumentos salariales y ceses laborales.
- Expertos en recursos humanos advierten que la IA carece del contexto y la empatía necesarios para evaluar la contribución real de una persona, convirtiendo patrones de datos en sentencias.
- El 43% de los directivos que evaluaron si la automatización podía sustituir un puesto ya han eliminado ese rol, transformando el debate teórico en desplazamiento real.
- La frontera entre IA como herramienta de apoyo y como sustituta del criterio humano se difumina, mientras el lugar de trabajo se reconfigura desde adentro y en silencio.
Dos tercios de los directivos estadounidenses consultan inteligencia artificial para decidir quién debe ser despedido. Así lo revela una encuesta de Resume Builder realizada a 1.342 responsables de empresa, un dato que condensa en un número la transformación silenciosa que atraviesa el mundo laboral. El 78% de estos gestores afirma que la IA determina quién recibe un aumento, el 77% que decide los ascensos, y el 66% que la emplea para los despidos. Las herramientas más utilizadas son ChatGPT, Copilot de Microsoft y Gemini de Google.
Lo que más preocupa a los expertos no es el uso de la IA como apoyo, sino los casos en que se convierte en árbitro definitivo. Aproximadamente el 2% de los directivos permite que el algoritmo tome la decisión final sobre despidos o promociones sin ninguna revisión humana. Stacie Haller, asesora de carrera en Resume Builder, advierte que la tecnología puede procesar datos, pero no comprende el contexto, la empatía ni la complejidad de cada situación personal.
El impacto ya es tangible: el 43% de los directivos que evaluaron si la automatización podía reemplazar un puesto de trabajo terminaron eliminándolo. No es una amenaza futura, sino un desplazamiento en curso. Al mismo tiempo, el 97% de los gestores usa la IA para crear materiales de formación y el 94% para diseñar planes de desarrollo profesional, manteniendo en esos casos la supervisión humana.
Lo que emerge es un lugar de trabajo en transición, donde la línea entre herramienta y sustituto se vuelve cada vez más borrosa. Cientos de miles de trabajadores se enfrentan a la posibilidad de que un sistema entrenado en patrones de datos, pero ajeno a su historia individual, decida si conservan su empleo.
Two-thirds of American managers are now asking artificial intelligence which employees to fire. That's the finding from a survey of 1,342 company managers conducted by Resume Builder, a employment services platform. The number lands like a punch: 66 percent of these decision-makers consult AI when deciding who should lose their job.
The practice sits at the center of a larger anxiety about what AI will do to work itself. Some analysts project that as many as 300 million jobs globally could be affected by automation in the coming years. Others argue the technology will simply handle repetitive tasks, freeing humans for more complex work. The European Central Bank has suggested that warnings about AI destroying employment may be overblown. Yet at the same time, an AI startup called Artisan ran an advertising campaign with the blunt slogan "Stop Hiring Humans," promoting AI agents as replacements for customer service and sales roles. The contradiction is stark: some see AI as a tool to augment human work, others as a tool to eliminate it.
But the Resume Builder survey reveals something more immediate than theoretical debate. It shows that managers are already delegating consequential decisions about people's livelihoods to machines. Sixty percent of the surveyed managers said they trust AI to make human resources decisions broadly. When the questions get more specific, the numbers climb: 78 percent say AI determines which employees get raises, 77 percent say it decides who gets promoted, and 66 percent use it to decide who gets fired. The tools they rely on are familiar ones—ChatGPT (53 percent), Microsoft's Copilot (29 percent), and Google's Gemini (16 percent).
What makes this more troubling is that roughly 2 percent of managers allow AI to make the final call on terminations or promotions without any human review. No second opinion. No conversation. Just the algorithm's recommendation becoming the decision. Stacie Haller, a senior career advisor at Resume Builder, warned against this approach. She noted that while AI can process data, it operates without the context, empathy, or judgment that personnel decisions demand. The technology simply cannot understand the full picture of a person's situation or contribution.
The survey also found that 46 percent of managers using AI for team management have been asked to evaluate whether the technology could replace an employee's role entirely. Of those, 57 percent concluded that automation was feasible, and 43 percent actually went ahead and eliminated the position. That's real displacement—not hypothetical, not future, but already happening.
Yet the picture is not uniformly grim. Many managers treat AI as a genuine support tool rather than a replacement for judgment. Ninety-seven percent use it to create training materials, 94 percent to design professional development plans, and 91 percent to set performance evaluation criteria. In these cases, managers review what the AI produces before implementing it. They maintain human oversight. The difference is crucial: AI as an aid to human decision-making versus AI as the decision-maker itself.
What emerges from the data is a workplace in transition, one where the line between tool and replacement has begun to blur. Hundreds of thousands of American workers now face the possibility that an algorithm, trained on patterns in data but incapable of understanding their individual circumstances, will determine whether they keep their jobs. The technology is not taking work away in the way some feared—it is not eliminating entire industries overnight. Instead, it is quietly reshaping how the people who manage other people make their most consequential choices.
Citações Notáveis
AI can support data-based decisions but lacks the context, empathy, and judgment that personnel decisions require— Stacie Haller, senior career advisor at Resume Builder
A Conversa do Hearth Outra perspectiva sobre a história
So two-thirds of managers are using AI to decide who to fire. That seems like a massive shift. What's actually happening in those conversations between a manager and the AI?
They're asking it to analyze data—performance metrics, attendance, productivity scores—and recommend who should go. The AI looks at patterns and spits back a name or a list. It's efficient. It feels objective.
But the survey mentions that only 2 percent let the AI make the final decision alone. So most managers are still involved?
Yes, but involvement isn't the same as judgment. A manager might use the AI recommendation as cover for a decision they wanted to make anyway, or they might defer to it because it seems more rational than their gut. The real problem is that the AI has no idea why someone might be struggling—illness, family crisis, a bad quarter in an otherwise solid career.
The survey also says 43 percent of managers have actually replaced employee positions with AI. That's not just deciding who to fire—that's eliminating the job entirely.
Right. And that's where the distinction between tool and replacement collapses. If you're using AI to decide who to fire, you're still keeping the role. But if you're using it to decide that the role itself should be automated, you're making a different kind of choice.
What about the managers who use AI for training and development? That seems more benign.
It is, because they're supervising the output. They review what the AI suggests before they use it. That's the model that actually works—AI as a draft, not as a verdict. The problem is that termination decisions don't get that same scrutiny. They feel too urgent, too final.
So the real fear isn't that AI will take all the jobs. It's that it will make the decisions about which jobs disappear, and those decisions will be made without the context that only a human can provide.
Exactly. The technology doesn't understand a person. It understands a data point.