Despite widespread adoption, the promise of artificial intelligence delivering significant productivity gains and widespread benefits remains largely unfulfilled, according to a new study from the National Bureau of Economic Research (NBER). The research, surveying around 6,000 CEOs, CFOs, and other senior leaders across the US, UK, Germany, and Australia, reveals that over 89% of firms haven’t seen a positive impact on productivity after implementing AI tools over the last three years.
The study does indicate growing interest in AI, with 69% of firms currently utilizing AI tools, and further growth anticipated. However, the lack of demonstrable productivity improvements raises questions about the current state of AI implementation and its actual value proposition for businesses. Interestingly, despite the limited productivity gains, a substantial 90% of companies haven’t reported any major job losses as a result of AI adoption, and 63% anticipate this will remain the case.
Current AI Use Cases
The NBER report highlights the most common applications of AI within organizations. Text generation leads the way, utilized by 41% of firms, followed by data processing (30%) and visual/image content creation (both at 30%). While less prevalent, AI is also being applied to robotics (9%) and autonomous vehicles (3%), though usage in these areas increased by approximately 50% during 2025, suggesting growing investment in more complex applications.
The limited impact on productivity isn’t deterring companies from continued investment, but it is tempering expectations. While 25% of firms anticipate small productivity gains and 12% expect large gains, a significant 60% foresee no further impact at all. The outlook for employment is similarly cautious, with 63% expecting no change, and only 26% anticipating job reductions.
Researchers estimate that, over the next three years, AI adoption will boost productivity by an average of 1.4%, while reducing employment by around 0.7%. This modest projection underscores the gap between the initial hype surrounding AI and the reality of its current impact.
The Productivity Paradox and Communication Gaps
The findings align with broader observations about the challenges of realizing the full potential of AI in the workplace. A recent report from DHR Global, detailed in their Workforce Trends Report 2026, echoes this sentiment, noting that nearly 39% of employees have reported noticeable productivity gains from AI tools, particularly in Asia, and Europe. However, this report also points to a critical disconnect: a lack of clear communication from leadership regarding how AI impacts roles, skills, and career paths.
This communication gap is crucial. Productivity gains alone aren’t enough to foster employee confidence or engagement. Workers need to understand how AI integrates into their daily work and how it shapes future opportunities. Without this clarity, uncertainty prevails, potentially hindering the effective adoption and utilization of AI tools.
AI Intensifies Work, Doesn’t Reduce It
Further complicating the picture is research from Harvard Business Review, published , which suggests that AI doesn’t necessarily reduce workload; instead, it tends to intensify it. The study found that employees using AI tools often work at a faster pace, take on a broader scope of tasks, and extend their working hours. This suggests that AI may be augmenting capabilities rather than replacing tasks, leading to increased demands on workers.
The Risk of Misinformation and Errors
The DHR Global report also highlights a practical challenge: the potential for errors and misinformation generated by AI tools. One in five respondents encountered inaccuracies or misleading outputs, requiring manual correction and potentially slowing down project progress. This underscores the need for robust quality control measures and careful oversight when utilizing AI-generated content.
Uneven Adoption and Geographic Disparities
Adoption of AI isn’t uniform across geographies or industries. An Anthropic Economic Index report points to uneven adoption rates, with stronger uptake in Asia and Europe compared to North America. This suggests that regional factors, such as investment in infrastructure, regulatory environments, and workforce skills, play a significant role in determining the success of AI implementation.
Generative AI in the Enterprise: A Nascent Stage
While the NBER study focuses on overall AI adoption, a report from Menlo Ventures highlights the relatively early stage of generative AI within the enterprise. The report suggests that while interest is high, many organizations are still grappling with how to effectively integrate generative AI into their workflows and realize its potential benefits.
The NBER researchers acknowledge that their study provides a snapshot in time and that further research is needed to establish more reliable comparisons. They emphasize the importance of standardized methodologies in future surveys to ensure consistency and accuracy. For now, the data suggests that while AI adoption is growing, the promised revolution in productivity remains largely unrealized, and organizations must focus on clear communication, careful implementation, and robust quality control to maximize the value of these powerful tools.
