Star employees widen their lead
People often say AI evens things out: a junior with Gemini or ChatGPT performs almost like a senior. But...what if it were the opposite? Someone who knows a lot about their subject gets more out of the tool, because they know what to ask, notice when it is wrong and use it where it matters. Someone who knows little has nothing to judge what comes back against. This paper calls it the "Matthew effect of AI", after the Gospel verse that says to those who have, more will be given.
Employees with more experience and strategic use of automatic content-creation platforms improved their performance relative to the rest, widening the performance gap. Personal automatic-creation tools increased their bargaining power by allowing high-value results independent of the company, while enterprise support systems strengthened the firm's ability to capture value.
In the daily practice of organizations, the arrival of generative AI has stopped being a novelty and become a tool that separates employees who use it strategically from those who use it superficially. When knowledge "stars" combine deep experience with the ability to create content, code or analysis instantly, their productivity takes off and the performance gap widens, defying the expectation that technology would level the playing field. This phenomenon, which the authors of the paper in *Human Resource Management* call the "AI-specific Matthew Effect", means that AI adoption does not only transform processes, it reshapes how value is distributed inside the company.
The central argument is that AI is not a simple efficiency multiplier but an agent that intensifies talent stratification. Employees who adopt personal automatic-creation tools can generate high-value results without depending on the firm's infrastructure, which gives them greater bargaining power and a form of portable capital. By contrast, corporate AI systems act as isolating mechanisms that make it easier for the company to capture that value, reinforcing its control over the outputs generated internally. Traditional human capital theory therefore has to be extended to recognize the complementarity between human skills and artificial capabilities, and to explain how that interaction produces both productivity gains and equity tensions.
Managers need to rethink their talent management and technology investment policies. First, it is essential to create governance frameworks that regulate access to personal AI tools, so they do not become weapons of disproportionate advantage. Second, incentives should be aligned with shared value creation, rewarding not only individual output but also collaboration that strengthens the company's resources. Finally, ongoing training should aim to raise the competence level of the whole workforce, so that the gap does not become a structural barrier but an opportunity to distribute the benefits of AI more equitably.
Conclusions
- Adopting automatic-creation platforms tends to widen performance differences between star employees and the rest.
- Personal automatic-creation tools increase workers' bargaining power by making their output more portable.
- Enterprise support systems act as isolating mechanisms that strengthen the firm's capture of value.
- Organizations need to rethink their human capital models to balance productivity and equity.