Automation now reaches the people who trained it.
Dublin, May 2026. Meta is eliminating more than 700 jobs in Ireland linked to content review and artificial intelligence training, exposing one of the sharpest contradictions in the current AI race. The workers affected helped supervise, classify and refine the systems that now allow the company to automate more of its operations.
The cuts form part of a broader restructuring as Meta intensifies its investment in artificial intelligence while reducing operational costs. The company is moving toward a leaner model in which automated systems absorb tasks once performed by human moderators, data reviewers and AI training teams. What disappears is not only employment, but a human checkpoint inside the platform’s decision-making machinery.
The case is especially sensitive because these roles were not peripheral to the digital ecosystem. Content reviewers and AI trainers operate in the hidden infrastructure of platforms, where harmful material is filtered, model behavior is corrected and system outputs are evaluated. Their work is often invisible to users, but essential to the quality, safety and legitimacy of automated products.
Meta’s decision reflects a broader pattern across the technology sector. Companies are spending heavily on AI infrastructure while cutting labor in areas they believe automation can replace or compress. The promise is efficiency; the risk is that fewer human workers remain inside systems that still require judgment, context and ethical supervision.
The deeper issue is not whether AI can reduce costs, but what happens when companies remove the people who understand the failures of those systems from the inside. Automation can scale decisions, but it can also scale errors, bias and opacity. When the human layer is weakened, platform governance becomes faster but not necessarily wiser.
Meta’s layoffs are therefore more than a corporate adjustment. They mark a turning point in the labor politics of artificial intelligence: the same workers who helped build and train the new system are now among the first to be displaced by it.
Detrás de cada dato, la intención. / Behind every data point, the intention.