Productivity, typically measured as output per hour worked, is considered the key long-term driver of higher income and better living standards. However, in both the USA and Europe, productivity growth since the mid-2000s has been weaker than in previous decades.
Many economists and policymakers now see artificial intelligence (AI) as a potential catalyst to counteract this downward trend. AI, particularly the increasing use of generative AI and AI agents, is expected to shape the next phase of productivity growth in developed countries such as the United States and countries in Europe.
The key question for managers is not whether AI will play a central role, but rather how large the productivity increases will be, how quickly they will occur and which regions will benefit most from them.
Increase in productivity through AI
The Organization for Economic Cooperation and Development (OECD) assumes that AI could increase labor productivity in industrialized countries by around 0.4 to 1.3 percentage points annually, depending on how intensively and across industries AI is used. These increases would be significant because even an additional half percentage point of annual growth over a decade has a significant overall effect.
However, the OECD and other economists emphasize that the results depend not only on technology, but above all on additional investments in digital infrastructure, workforce training and organizational changes.
Between 1995 and 2019, U.S. labor productivity rose 2.1 percent annually, compared to 1 percent in Europe. This discrepancy is due in part to the fact that companies in the United States invested more aggressively in information and communications technology and other technologies, while in Europe they were subject to greater regulatory requirements and other factors.
When it comes to productivity gains through the use of AI, expectations in the USA remain higher than in Europe. Analyst firm estimates Goldman Sachs According to this, labor productivity in the USA could increase by around 1 to 1.5 percentage points per year through the use of generative AI.
This forecast is based on several structural factors: On the one hand, the USA has a comprehensive technology ecosystem and is a global pioneer in AI research and venture capital. On the other hand, there is a large service sector in which digitalization is already well advanced. Generative AI tools can be used quickly here, for example in finance, consulting and the IT sector.
Agent-based AI
In both Europe and the US, AI agents represent a particularly important trend. Unlike older automation tools that handle separate tasks, AI agents – for example SAP's Joule Agents – are designed to plan, reason and execute multi-step workflows. For example, an agent can coordinate customer reports, suggest responses, run queries on databases, escalate problems, and update systems. Human intervention is only necessary to a certain extent.
In knowledge-based industries, this type of workflow automation could significantly improve the performance of individual employees. However, rather than replacing entire professions, AI agents could reduce the time spent on repetitive administrative activities and tedious tasks. Employees would be able to concentrate on value-adding analyzes and strategies as well as on activities that require communication skills.
Despite reports of failed corporate AI projects, which were typically back-of-the-envelope or standalone AI pilots rather than an integrated, holistic approach, recent evidence from the US suggests that productivity gains are already emerging in some industries. Financial institutions, for example, have seen significant increases in efficiency in back-office processes through the use of AI.
Experimental studies in the area of consulting also show that generative AI can increase quality and speed, especially for less experienced employees. In this way, skills gaps within teams can be effectively closed.
Prospects for growth in Europe thanks to AI
The prospects for productivity increases through AI in Europe are rather mixed. A current one International Monetary Fund (IMF) report According to this, medium-term productivity gains from AI alone would vary significantly depending on the country. For Europe as a whole, the overall increase would be rather modest at around 1.1 percent over a five-year period.
However, the IMF suggests that significantly larger increases are possible in the long term with growth-enhancing reform measures. Like the OECD, the IMF emphasizes that legal frameworks, labor market structures and the pace of technology diffusion will have a major impact on outcomes.
Several structural differences shape Europe's trajectory and the extent of financial benefits from the growth of AI. First, AI adoption tends to be slower among small and medium-sized businesses, which make up a larger share of the economy in Europe than in the US. Second, Europe's digital market is more fragmented across national borders, languages and legal systems, which can make it difficult to scale up technology platforms. And third, the European Union is taking a more cautious approach to regulating AI. While this can reduce certain risks, it can also dampen short-term productivity gains if regulatory requirements slow the adoption of AI.
The strengths of Europe: AI agents embedded in industrial systems
But Europe also has its strengths and is, for example, a leader in modern manufacturing and industrial technology. In these areas, AI-supported optimization measures, robotics and predictive maintenance could increase capital productivity. AI agents embedded in industrial systems could significantly improve supply chain efficiency and reduce downtime.
As SAP executives have pointed out, Europe also has an enormous volume of structured business and manufacturing data, which is essential for reliable and effective AI systems and trust in AI agents.
As the adoption of AI in manufacturing and energy systems accelerates and European companies seize the opportunity to develop sophisticated AI agents and apps using their business data, Europe could see significantly higher productivity gains in the medium term. For example, through the internal use of AI tools, SAP has already been able to significantly increase the productivity of its own development departments.
Flexible labor market
The adaptability of the labor market is a crucial factor in both the USA and Europe. Historically, the U.S. labor market has been more flexible, characterized by more frequent job changes and greater job mobility. This flexibility allows employees to be moved more quickly into AI-supported roles, which can increase productivity. More effective retraining of the workforce could further increase productivity.
As the Bank for International Settlements (BIS) noted, the productivity effects of AI are unlikely to occur automatically. Productivity gains from AI depend on additional investments in employee skills, management practices and digital infrastructure. The BIS also warns that without these investments, AI tools may only deliver modest efficiencies.
Experience with previous cross-cutting technologies such as electricity and IT shows that productivity gains only occur when companies redesign processes to exploit new opportunities and take a holistic approach to implementation rather than a fragmentary approach.
AI is not a bubble
While some investors have expressed concerns about an AI bubble, total spending on AI in the US is still less than 1 percent of gross domestic product (GDP). According to Joseph Briggs, senior global economist at Goldman Sachs, spending is well below historical investment cycles. For comparison: in the past, investments in infrastructure such as IT, railways and shipping canals were typically 2 to 5 percent of GDP.
Similar to these past investment efforts, AI – particularly agent-based AI – is expected to deliver significant productivity growth and a corresponding increase in GDP in the regions and industries that fully realize the potential of AI.
