New study reveals generative AI boosts job growth and productivity

A new study analyzing more than a decade of U.S. patent data has found that not all artificial intelligence (AI) innovations displace human workers—some, particularly generative AI, actually augment jobs and drive firm growth.

The research, posted on the SSRN preprint server and titled “Displacement or Augmentation? The Effects of AI Innovation on Workforce Dynamics and Firm Value,” was conducted by Mark Chen, professor of finance at Georgia State’s J. Mack Robinson College of Business and Joanna (Xiayou) Wang at Peking University’s HBSC Business School and is one of the first large-scale, empirical efforts to disentangle the nuanced effects of AI technologies on labor and firm performance.

The researchers set out to answer the question: Do AI technologies complement or substitute human workers, and what are the implications of these effects for firm productivity and value?

In analyzing more than 5 million U.S. patents from 2007 to 2023, the researchers examined seven different AI capabilities that mirror human intelligence and how these technologies affect employment and business productivity, shedding light on which AI technologies enhance human labor and which may replace it.

Among the key findings is that generative AI is an engine for job growth and innovation. AI technologies classified under language, learning, creativity, engagement, and decision-making functions—such as language models, content creation tools, and intelligent recommendation systems—were shown to augment human workers, leading to more hiring, greater productivity, and higher firm value. Occupations that tend to be augmented by AI are most white-collar occupations like finance, management, and science and engineering, as well as occupations in art, design, and entertainment.

The study also found that when AI can displace workers, it cuts costs but doesn’t increase productivity. In contrast, AI used for perception and motor control tasks—like image and speech recognition—tends to replace human labor, particularly in routine or repetitive roles. While this reduces labor costs to a firm, it does not significantly improve firm productivity. The type of occupations that are being displaced by AI are manual labor jobs, like farming, fishing, health care support, and building and grounds cleaning/maintenance.

In addition, labor market flexibility determines how valuable AI is for a firm. When a firm can easily hire skilled workers, they gain more from augmenting AI. Conversely, displacing AI is more effective when firms can terminate employees with fewer constraints, such as low severance costs or strict non-compete laws.

“We found the impact of AI on employment and firm value is very nuanced—it isn’t going to just destroy jobs, but it’s also not going to always create new jobs. Many types of generative AI lead to higher employment of workers who bring new skills to the workplace and to higher headcount at a firm,” said Chen. “Companies that invest in these types of AI tools often become more productive by complementing their workforce, not replacing it.

“But the value of augmenting AI can be stifled if a firm has less flexibility in external hiring, like in states with higher unionization. And where firms could save on wage costs from displacing jobs with AI, those cost savings could be reduced in states with high unemployment insurance benefits because workers in those states are already paid lower wages.”

The study underscores the urgent need for reskilling programs and labor policies that help workers transition into AI-augmented roles. It also provides critical guidance for business leaders and policymakers navigating the AI revolution.