New York, NY – As artificial intelligence rapidly reshapes the American workplace, corporate leaders are inaccurately assuming what actually drives successful AI adoption. While most believe that success with AI adoption comes down to technology investment, company size, or industry, a survey from Columbia Business School finds that the single greatest predictor of whether a company successfully adopts and deploys AI across its organization has to do with how well its leaders listen to, support, and involve their employees. Organizations that do so are seven times more likely to succeed.
The survey, Employee Centricity in an AI World, was conducted by Columbia Business School Professor Stephan Meier with researchers at the BCG Henderson Institute and is based on responses from approximately 1,400 U.S. employees and leaders across industries, seniority levels, and company sizes. They found that how well a company treats its employees predicts AI success more than industry, department, or company size. In fact, when researchers looked at what explains why some companies are better at AI than others, employee centricity was the single biggest factor. The survey also found a sharp emotional divide: 33% of frontline employees list more negative than positive emotions around AI, compared to just 4% of executives, with the most common negative emotions being anxiety, resistance, and fear of job loss. This aligns with a broader perception gap, namely that while 80% of executives believe their employees feel well-informed about their organization's AI strategy, only 30% of employees agree. Yet the research also points to a clear path forward: at employee-centric organizations, workers are 70% more likely to feel optimistic about AI and 36% more likely to remain with their employer one year later.
The findings carry significant implications for how organizations approach workforce wellbeing in the age of AI, suggesting that employee mental health and organizational AI success are not separate challenges but deeply interconnected ones. Meier and his co-authors point to several concrete steps organizations can take to close the gap, including listening continuously to employee perspectives and measuring their experience regularly rather than once a year; communicating transparently not just how AI will be deployed but why; and involving employees directly in shaping how AI tools are introduced. The survey suggests that for organizations looking to get AI right, the place to start is with their people.
