Mehdi Paryavi, CEO of the International Data Center Authority, warns that overreliance on AI in the workplace could be silently undermining workers’ confidence and long-term career potential. While AI is often praised for boosting speed, efficiency, and output, Paryavi argues it may also be triggering a subtle but dangerous decline in critical thinking and self-assurance. Speaking with Business Insider, Paryavi described a growing phenomenon he calls "quiet cognitive erosion" — a gradual weakening of human skills due to excessive dependence on AI. He noted that the once-valued ability to "think outside the box" is at risk of disappearing, as employees increasingly draw their ideas, analysis, and creativity from a single source: AI. According to Paryavi, confidence is the first casualty. "If you start believing that AI writes better than you and thinks smarter than you, you’ll lose faith in your own abilities," he said. Over time, this erosion compounds as workers begin delegating core tasks—writing, analysis, decision-making—to AI, reducing their own engagement and skill development. This shift is already showing signs in the workforce. A recent report from the Work AI Institute, developed with researchers from institutions including Harvard, Notre Dame, and UC Santa Barbara, found that while AI makes employees feel more productive and competent, it may simultaneously weaken their foundational skills. Rebecca Hinds, head of the institute, explained that AI creates an illusion of expertise—particularly dangerous for early-career professionals still building their capabilities. Paryavi stresses that speed is not the same as true productivity. While AI can generate polished reports and rapid insights, it often lacks the depth, context, and nuanced understanding that come from years of experience and independent thought. Anastasia Berg, a philosophy professor at the University of California, Irvine, has warned that heavy AI use can lead to rapid skill atrophy, especially among junior staff who never fully develop the ability to solve problems on their own. The solution, Paryavi says, isn’t to reject AI but to use it thoughtfully. He recommends that companies tailor AI access based on job function rather than applying it universally. Some roles may benefit from AI assistance, while others should prioritize human judgment and creativity. He also emphasizes the importance of human involvement at both ends of the workflow: leading with original thinking at the start and rigorously reviewing AI output at the end. "You must quality check AI—not the other way around," he said. AI may not eliminate jobs, but without intentional boundaries, Paryavi warns, it could quietly erode the very skills and confidence that define professional growth. "How much technology do we really need?" he asked. "And how far are we willing to go before we lose ourselves in the tools we built?"