How AI is affecting productivity and jobs in Europe

AI boosts European productivity by 4 percent

Artificial intelligence adoption increases labor productivity in the European Union by an average of 4 percent without reducing overall employment. A new study from the Bank for International Settlements and the European Investment Bank analyzed 12,000 European firms to find the first causal evidence of how the technology affects the continent's economy. While the gains are significant, they fall short of the most optimistic predictions of a total economic transformation.

Researchers used a novel strategy to identify these effects by matching European firms with comparable US companies. They looked at sector, size, investment intensity, and management practices to isolate the specific impact of AI from other regional economic factors. This method confirmed that AI acts as a complementary tool that helps workers perform tasks more efficiently rather than replacing them entirely.

The 4 percent productivity boost represents a short-term gain in efficiency through capital deepening. AI allows employees to complete professional writing, coding, and data analysis tasks faster and with higher accuracy. These findings align with recent experimental data showing that specific AI tools can increase individual task productivity by 10 to 65 percent.

Adoption rates vary across the continent

European AI adoption currently mirrors levels seen in the United States, but the data hides a significant internal divide. Financially developed countries like Sweden and the Netherlands show high adoption rates, with 36 percent of firms using big data and AI in 2024. These nations effectively match the technological pace of the American market.

Less financially developed economies in the EU lag behind their northern neighbors. Adoption rates in Romania and Bulgaria hover around 28 percent, and this gap has widened over the last few years. The disparity suggests that local financial infrastructure plays a massive role in whether a company can afford to integrate new technologies.

The EU continues to trail both the US and China in AI innovation metrics despite these adoption numbers. The US leads in the absolute number of AI patents and specialized AI research. Europe’s AI preparedness index, which measures digital infrastructure and human capital, shows that many member states remain unequipped to lead the next wave of development.

Large firms see the biggest gains

Company size remains the strongest predictor of whether a business will successfully deploy AI. Approximately 45 percent of large firms with more than 250 employees have integrated AI into their workflows. Only 24 percent of small firms with 10 to 49 employees have done the same.

Large companies possess the economies of scale and technical expertise required to handle high integration costs. They also have the data infrastructure necessary to train or implement large language models and analytics tools. This concentration of technology among the biggest players creates a risk of widening the gap between industry leaders and smaller competitors.

The study highlights several characteristics common among AI adopters:

• They invest more heavily in innovative research and development.
• They face more significant challenges in finding highly skilled workers.
• They maintain more flexible management structures that allow for workflow redesign.
• They have better access to external financing and capital markets.

AI does not destroy jobs

The data shows no evidence that AI adoption reduces the total number of employees at a firm in the short term. Earlier fears of mass displacement have not materialized in the European labor market. Instead, AI appears to augment worker output, allowing the existing workforce to handle more complex workloads.

Workers at firms using AI have actually seen higher aggregate wages. These employees are becoming more valuable as they learn to use AI to make better decisions and automate routine administrative tasks. The current trend points toward "capital deepening," where the technology makes each hour of human labor more productive.

This stability in employment numbers might be transitional as firms still learn how to use the technology. As AI systems become more capable, the risk of labor displacement could increase. For now, the primary effect is an increase in efficiency and wage growth for the people already in these roles.

Software and training multiply AI results

Simply buying an AI license is not enough to see a productivity return. The study found that complementary investments in intangible assets are the real drivers of success. Companies that invest in their own data infrastructure and staff training see much higher returns on their AI spending.

The numbers reveal a clear multiplier effect for firms that pair AI with other investments:

• Every 1 percentage point increase in software and data spending boosts AI productivity by 2.4 percentage points.
• Every 1 percentage point increase in employee training boosts AI productivity by 5.9 percentage points.
• Firms with modern management practices see double the productivity gains compared to traditional firms.

These findings suggest that "fusion skills" are the most valuable asset in the current economy. These include prompt engineering, data stewardship, and human-in-the-loop decision making. Companies that prioritize these skills see the 4 percent productivity gain jump significantly higher.

Policy must focus on small firms

European policymakers need to address the scale problem to prevent a permanent productivity divide. Since the benefits of AI are currently concentrated in medium and large firms, smaller companies need help reaching a critical mass. This requires deeper financial markets that can provide capital to innovative startups and SMEs.

Advancing the EU Savings and Investment Union is a necessary step to ensure smaller firms can compete. Public policy should move beyond just subsidizing hardware and software. Governments should instead incentivize workflow redesign and continuous learning programs that help firms integrate technology into their daily operations.

While the short-term data is encouraging for workers, long-term monitoring is required. The current wage gains may skew toward highly skilled workers, potentially increasing income inequality across the continent. Proactive education policies and vocational training will be essential to ensure the 4 percent productivity boost benefits the entire workforce.