Not long ago, artificial intelligence was being celebrated as the future of business. Enthusiasts and executives alike predicted that AI would handle routine tasks, boost efficiency, and reduce staffing costs.
OpenAI’s Sam Altman once called 2025 the breakthrough year for “AI agents”—digital workers meant to perform complex tasks independently. But now, midway through the year, those promises are starting to unravel.
AI Agents: More Hype Than Help?
The idea behind AI agents was simple: create digital employees that operate around the clock without the limitations of human needs. These agents were expected to complete full tasks autonomously, ideally without oversight.
But the reality hasn’t matched the hype. By April 2025, even the most capable AI agents could only handle about 24% of assigned tasks. That’s a far cry from expectations. Nonetheless, many companies jumped the gun—laying off teams and replacing skilled workers with AI systems that haven’t delivered.
What some viewed as innovation is now looking more like poor judgment.
Executive Doubts Are Growing
According to a recent Gartner survey of 163 executives, half said they’re abandoning earlier plans to drastically reduce customer service staff by 2027.
The reason? AI tools are falling short—especially in customer service, where understanding nuance, escalating issues correctly, and interpreting intent remain major challenges. The shortcomings have forced companies to rethink their approach.
Terms like “AI-powered transformation” are fading from corporate language. In their place: buzzwords like “hybrid workforce,” “transitional hurdles,” and “human-in-the-loop”—a clear acknowledgment that human workers remain essential.
As Kathy Ross, a senior director at Gartner, noted: “The human touch remains irreplaceable in many interactions.” In the end, it seems emotional intelligence can’t be coded.
