You put these technologies into old social structures, and everything creates friction.
Entre marzo de 2025 y marzo de 2026, más de 107,000 trabajadores en el mundo perdieron su empleo en recortes vinculados directamente a la inteligencia artificial, una cifra que los analistas consideran apenas el umbral de una transformación estructural más profunda. Gigantes como Intel, Amazon y Microsoft no ajustaron sus plantillas por ciclos económicos, sino por una reconfiguración deliberada del trabajo humano. Lo que emerge no es solo una estadística laboral, sino una pregunta civilizatoria: ¿a qué velocidad pueden las sociedades adaptarse cuando la tecnología avanza más rápido que las instituciones que la contienen?
- La cadencia de despidos en el sector tecnológico global alcanzó aproximadamente 900 trabajadores diarios en 2026, una aceleración que convierte cada jornada laboral en una cuenta regresiva para miles de personas.
- Las reestructuraciones de Intel, Amazon y Microsoft no se justificaron por pérdidas financieras, sino por la promesa de que la IA podía ejecutar tareas de ingeniería y administración mejor y más barato que los humanos.
- En México, la amenaza es más silenciosa: las empresas no despiden masivamente, pero dejan de contratar, y solo el 18.6% de las compañías usa IA activamente, lo que ofrece una protección temporal pero no permanente.
- Especialistas advierten que la verdadera vulnerabilidad no es tecnológica sino humana: la falta de alfabetización digital y de formación ética expone a los trabajadores a quedar desplazados sin herramientas para reintegrarse.
- Los expertos proyectan que la integración estable de la IA en los mercados laborales podría requerir entre dos y tres décadas, un horizonte que convierte los 107,000 despidos actuales en apenas el primer capítulo de una historia mucho más larga.
Entre marzo de 2025 y marzo de 2026, al menos 107,756 personas perdieron su empleo en recortes tecnológicos atribuidos directamente a la inteligencia artificial. La ola comenzó cuando Siemens anunció la eliminación de seis mil puestos para integrar sistemas de automatización, y siete grandes corporaciones siguieron su ejemplo en rápida sucesión. Intel recortó 48,900 trabajadores; Amazon, 30,000; Microsoft, 15,100. No eran ajustes menores: eran reorganizaciones estructurales justificadas por la capacidad de la IA para sustituir tareas que antes realizaban personas. Para 2026, el ritmo había escalado a cerca de 900 despidos diarios en el sector tecnológico global.
Esteve Almirall, autor de un libro sobre el impacto de la IA, estimó que la cifra real podría superar los 150,000 afectados si se contabilizan todas las grandes corporaciones. Su diagnóstico, sin embargo, apuntaba más allá de los números: el problema central no era tecnológico sino social. Las organizaciones y los modelos laborales no evolucionaban al ritmo de la innovación, generando una fricción que, advirtió, podría tardar décadas en resolverse.
México ofrecía un panorama distinto, aunque no tranquilizador. Fernando Sentíes, director de AMITAI, señaló que en el país no se observaban reemplazos masivos de trabajadores por agentes de IA, sino algo más sutil: los puestos administrativos no se cubrían cuando quedaban vacantes y las rotaciones de contratación se ralentizaban. Solo el 18.6% de las empresas mexicanas usaba inteligencia artificial de forma activa, y la mayoría carecía de la capacidad organizacional para transformar sus procesos de manera acelerada.
Sin embargo, la especialista Ivete Sánchez Bravo advirtió que esa relativa inmunidad era temporal. El impacto ya era visible en las multinacionales que operan en México y en las empresas mexicanas que compiten a escala global. La verdadera exposición, subrayó, residía en la falta de formación digital y ética de los trabajadores: quienes no desarrollen competencias para colaborar con sistemas de IA quedarán vulnerables. Los 107,000 despidos documentados, concluyeron los expertos, son solo el inicio visible de una transformación que apenas comienza a desplegarse.
Something shifted in the global economy between March 2025 and March 2026. Not gradually, but with a visible acceleration—the kind that makes analysts sit up and start counting. When researchers at El Universal tallied the numbers across major technology companies, they found that at least 107,756 people had lost their jobs in cuts explicitly tied to artificial intelligence. The wave began in mid-March 2025 when Siemens, the German industrial giant, announced it was eliminating six thousand positions worldwide to integrate AI systems that would automate code writing and engineering tasks. Seven other major corporations followed in quick succession.
The scale became apparent only when the full roster emerged. Intel shed 48,900 workers. Amazon cut 30,000. Microsoft eliminated 15,100. Meta removed 1,300 from its payroll. These were not small adjustments or seasonal fluctuations—they were structural reorganizations justified by the promise of artificial intelligence to do work that humans had been doing. By 2026, the pace had accelerated to roughly 900 dismissals per day across the technology sector globally. The specialized tracking site Layoffs.fyi documented that since 2020, the tech industry had accumulated more than 450,000 job cuts worldwide, but the character of those cuts was changing. Earlier reductions had been blamed on pandemic-era overhiring. Now, increasingly, they were being tied directly to AI adoption.
Esteve Almirall, author of "When Everything Changes AI," offered a sobering perspective in conversation with El Universal. He estimated that the true figure might reach 150,000 when accounting for all major corporations. But his concern extended beyond the numbers themselves. The real problem, he argued, was not technological but social. "You put these technologies into old social structures," he explained, "and everything creates friction." Organizations and labor models were not evolving at the pace of innovation. The gap between what machines could do and what societies were prepared to accept created a grinding tension that could take decades to resolve.
Mexico presented a different picture, at least for now. Fernando Sentíes, CEO of AMITAI, the Mexican Association of the Information Technology Industry, cautioned against assuming that what was happening in Silicon Valley would immediately replicate south of the border. He had not observed widespread replacement of human workers with AI bots or intelligent agents in Mexican companies. What he had seen was more subtle: administrative positions were not being refilled when people left, and hiring rotations were slowing. But he emphasized a crucial distinction. "It's very difficult for artificial intelligence right now to replace positions that require human judgment," he said. Moreover, not all job losses in Mexico could be attributed to automation. When the country lost 25,000 employers in the social security system, those losses rippled through the labor market in ways that had nothing to do with AI.
Sentíes noted that only 18.6 percent of Mexican companies were actively using artificial intelligence, and that figure did not mean they were replacing workers at that rate. Mexican firms faced internal constraints that prevented rapid transformation. "They're not ready yet, even though the technology is, to change all their processes overnight," he observed. The adoption curve in Mexico was slower, more deliberate, constrained by organizational capacity and capital investment. Yet Ivete Sánchez Bravo, another specialist tracking these trends, warned that Mexico's relative immunity was temporary. The impact was already visible in large multinational corporations operating in the country and in Mexican firms competing at a global scale. She called it an "inevitable trend" that would intensify, particularly in large corporate environments.
The vulnerability, Sánchez Bravo emphasized, was not primarily technological but human. Workers lacked digital literacy and ethical training in how to work alongside AI systems. That gap created exposure: failing to upskill could lead directly to job loss. She also stressed the importance of understanding what these tools actually were. Chatbots, for instance, were not designed to tell the truth. Used without critical judgment, they could spread misinformation. The real challenge was not the machines themselves but how societies would integrate them into existing structures.
Almirall projected that this transition period would be both crucial and turbulent. He estimated that humanity might need two to three decades to achieve stable integration of artificial intelligence into daily life and labor markets. During that span, he warned, tensions would persist, inequalities would deepen, and employment itself would be reconfigured in ways not yet fully visible. The 107,000 workers who had already lost their jobs were the early casualties of a transformation that was only beginning to unfold.
Citas Notables
The problem is that you put these technologies into old social structures, and therefore everything creates friction. Organizations and labor models do not evolve at the same pace as innovation.— Esteve Almirall, author of 'When Everything Changes AI'
I have not seen many Mexican companies substituting humans with bots or AI agents yet. What is happening is more subtle—administrative positions are not being renewed, but it is very difficult for AI right now to replace positions requiring human judgment.— Fernando Sentíes, CEO of AMITAI
La Conversación del Hearth Otra perspectiva de la historia
Why does it matter that these layoffs are explicitly tied to AI, rather than just being another round of corporate downsizing?
Because it signals a shift in how companies justify cuts. Before, they blamed the pandemic or market cycles. Now they're saying the work itself is being replaced by machines. That's a different conversation—it's not temporary, it's structural.
But the numbers seem enormous. Over 100,000 in a year. Is that actually unusual?
Not in isolation. The tech sector has cut 450,000 jobs since 2020. But the acceleration is real—900 people a day in 2026 alone. And the speed is what matters. These aren't gradual transitions; they're sudden reorganizations.
Mexico seems to be moving slower. Is that a good thing?
It's complicated. Slower adoption means fewer immediate job losses, which sounds good. But it also means Mexican workers and companies are less prepared when the wave does arrive. And it will arrive—especially in multinational firms already operating there.
What's the actual risk to a worker in Mexico right now?
The immediate risk is small if you work in a small or medium firm. But if you're in a large corporation or competing globally, the risk is real and growing. And the deeper risk is that most workers don't have the digital skills to adapt when it does happen.
Almirall said this could take 40 to 80 years to fully play out. That seems impossibly long.
He's talking about when societies fully absorb and reorganize around the technology. The job losses will happen much faster. The friction he's describing—old institutions meeting new tools—that's what creates the long timeline.
So what should someone actually do about this?
That's the question no one has a good answer to yet. Upskill, certainly. But in what? The targets keep moving. The real answer is probably systemic—education, policy, social safety nets—not individual.