Most companies have yet to see any productivity gains from artificial intelligence (AI) despite its widespread adoption, according to a comprehensive survey conducted by the US National Bureau of Economic Research (NBER). The study, which surveyed 6,000 executives from various industries across the US, UK, Germany, and Australia, revealed that while 70% of firms are actively using AI, a significant portion of top executives—including CFOs and CEOs—remain disengaged. Specifically, one-quarter of these executives reported not using AI at all, and two-thirds indicated they only use the technology for a maximum of 1.5 hours per week.
Over the past three years, nearly 90% of companies stated that their AI usage had no measurable impact on either employment or productivity. This finding aligns with prior research from the Massachusetts Institute of Technology (MIT), which highlighted that approximately 95% of AI pilot programs failed to generate productivity improvements, as well as a study from the Danish National Bureau of Economic Research that observed minimal effects on productivity.
The NBER survey indicated a varied picture of AI integration across the surveyed nations, with 69% of businesses employing the technology in some capacity. The US led with 78%, followed by the UK at 71%, Germany at 65%, and Australia at 59%. Respondents expect this figure to rise to 75% within the next three years. The findings also pointed out that larger, more productive companies, particularly those that offer higher wages, are more likely to utilize AI compared to older firms and those with senior management lacking technological acumen.
The most common applications of AI included “text generation using large language models,” “visual content creation,” and “data processing using machine learning.” Despite the potential of these technologies, 90% of executives reported no observable impact on employment over the past three years, while 89% noted a lack of productivity gains. The report acknowledged that while specific tasks, such as customer support and writing, have shown substantial productivity improvements, overall economic gains attributable to AI remain uncertain, with estimates suggesting a total factor productivity increase of only 1.5 percentage points over a decade.
In light of the current findings, business leaders remain optimistic about the future. Despite the lack of significant impacts thus far, executives predict that AI will begin to enhance productivity and reduce workforce requirements in the coming years. Specifically, respondents forecast a 1.4% productivity boost and a 0.8% increase in output due to AI advancements over the next three years. However, the survey also exposed a stark contrast in expectations between senior executives and staff; the former anticipate a 0.7% decrease in employment, equating to 1.75 million fewer jobs, while staff predict a 0.5% rise in employment driven by AI.
The NBER study arrives on the heels of comments from Microsoft AI CEO Mustafa Suleyman, who warned that “white-collar” occupations are likely to bear the brunt of AI’s impact in the near future. In a recent interview with the Financial Times, Suleyman stated that AI is expected to reach “human-level performance” in the next 12 to 18 months, potentially leading to widespread automation of professional roles such as lawyers, accountants, and project managers.
“I think we’re going to have human-level performance on most if not all professional tasks,” Suleyman remarked, emphasizing the rapid advancements in AI capabilities. As organizations increasingly embrace AI, the sentiment among executives suggests that while the technology has not yet delivered on its promises, the expectation is that its role will grow significantly, shaping the future of work in ways that are yet to be fully realized.
See also
AI Stock Volatility: Anthropic Launch Triggers Software Selloff Amid Market Uncertainty
Chelsea Secures Front-of-Shirt Sponsorship with IFS, Enhancing AI Partnership Until 2028
Swansea University Launches Google DeepMind-Supported AI Research Programme for Disadvantaged Students
ServiceNow Achieves $3.47B in Q4 Revenue, Doubles AI Product Sales Amid Market Sell-Off
Reddit Launches AI Shopping Search Tool Connecting Community Recommendations to Retailers
