How AI is affecting productivity and jobs in Europe

Hacker News · Feb 19, 2026 · Collected from RSS

Full Article

Europe faces a critical choice in the race for artificial intelligence (AI). As the technology promises to reshape economies worldwide, policymakers are caught between two competing narratives. Optimists envision AI as the catalyst for a new productivity boom, potentially adding several percentage points to annual growth (Baily et al. 2023). Sceptics warn that adoption barriers, skill gaps, and uneven diffusion may limit gains and exacerbate inequality (Acemoglu 2024, Filippucci et al. 2024, Gambacorta and Shreeti, 2025). For Europe, the stakes are particularly high: while the continent boasts world-leading AI researchers and industrial capacity, it lags behind the US and China in developing new AI technologies (Cornelli et al. 2023). Recent studies suggest that AI could widen cross-country income gaps, with benefits concentrating in advanced economies that are better prepared to adopt and integrate these technologies (Cazzaniga et al. 2024, Gambacorta et al. 2025, Hennig and Khan 2025). Yet robust firm-level evidence on AI’s actual effects in Europe remains scarce. Do European firms that adopt AI genuinely become more productive? Does AI destroy jobs or augment workers? Are the benefits shared broadly, or do they concentrate among larger, better-resourced companies? In a recent study (Aldasoro et al. 2026), we provide the first causal evidence on how AI adoption affects productivity and employment across more than 12,000 European firms. Europe’s AI paradox Europe’s position in the global AI landscape is paradoxical. On various innovation metrics, the continent falls behind. The EU trails the US not only in the absolute number of AI-related patents but also in AI specialisation – the share of AI patents relative to total patents. This innovation gap translates into differences in firms’ readiness to adopt AI, as measured by the IMF’s AI preparedness index, which assesses countries based on digital infrastructure, human capital, innovation capacity, and regulatory frameworks (Cazzaniga et al. 2024). However, when it comes to actual deployment, the picture is more nuanced. Drawing on the European Investment Bank Investment Survey (EIBIS), we find that on average, AI adoption levels are similar in the EU and the US. Notably, important heterogeneity emerges beneath the surface. Financially developed EU countries – such as Sweden and the Netherlands – match US adoption rates, with around 36% of firms using big data analytics and AI in 2024. In contrast, firms in less financially developed EU economies, such as Romania and Bulgaria, lag substantially behind, with adoption rates around 28% in 2024. Figure 1 illustrates this divide, showing how the gap has persisted and even widened in recent years. Figure 1 Use of big data analytics and AI by country groups Notes: Average share of firms reporting that they use AI by country groups, controlling for firms’ sector. The error bars represent 95% confidence intervals. EU countries are grouped based on an index of financial development using financial market data from 2015 to 2023 and consisting of two composite indicators: (i) financial market size and integration, and (ii) financial market depth (see Betz et al., 2026). Source: EIBIS 2019-2024. Adoption also varies dramatically by firm size. Among large firms (more than 250 employees), 45% have deployed AI, compared with only 24% of small firms (10 to 49 employees). This echoes classic patterns in technology diffusion (Comin and Hobijn 2010): larger firms possess the resources, technical expertise, and economies of scale needed to absorb integration costs. AI-adopting firms are also systematically different – they invest more, are more innovative, and face tighter constraints in finding skilled workers. These patterns suggest that simply observing which firms adopt AI and comparing their performance could yield misleading results, as adoption itself is endogenous to firm characteristics. Isolating AI’s causal effect To credibly identify the causal effect of AI on productivity, we develop a novel instrumental variable strategy, inspired by Rajan and Zingales’ (1998) seminal work on financial dependence and growth. Their key insight was that industry characteristics measured in one economy – where they are arguably less affected by local distortions – can serve as an exogenous source of variation when applied to other countries. We extend this logic to the firm level. For each EU firm in our sample, we identify comparable US firms – matched on sector, size, investment intensity, innovation activity, financing structure and management practices. We then assign the AI adoption rate of these matched US firms as a proxy for the EU firm’s exogenous exposure to AI. Because US firms operate under different institutional, regulatory and policy environments, their adoption patterns capture technological drivers that are plausibly independent of EU-specific factors. Rigorous propensity-score balancing tests confirm that our matched US and EU firms are virtually identical across key observable characteristics, validating the identification strategy. Our analysis draws on survey data from EIBIS combined with balance sheet data from Moody’s Orbis. Productivity gains without job losses Our results reveal three key findings. First, AI adoption causally increases labour productivity levels by 4% on average in the EU. This effect is statistically robust and economically meaningful, though more moderate than the transformative scenarios predicted by some observers. The magnitude aligns with mid-range macroeconomic projections (Bergeaud 2024) rather than the most optimistic estimates of productivity booms. While our analysis focuses on labour productivity levels and captures a one-off effect – rather than long-run total factor productivity growth – the 4% gain suggests that AI acts in the short term as a complementary input that enhances efficiency, albeit with implementation frictions and skill gaps tempering its impact. Second, and crucially, we find no evidence that AI reduces employment in the short run. While naïve comparisons suggest AI-adopting firms employ more workers, this relationship disappears once we account for selection effects through our instrumental variable approach. The absence of negative employment effects, combined with significant productivity gains, points to a specific mechanism: capital deepening. AI augments worker output – enabling employees to complete tasks faster and make better decisions – without displacing labour. This finding resonates with micro-level experimental evidence showing that AI tools can produce productivity gains between 10% and 65%, with strong effects in coding, consultant tasks and professional writing (Noy and Zhang 2023, Gambacorta et al. 2024, Brynjolfsson et al. 2025). These experimental effects are task-specific, whereas our estimates capture firm-level averages. Importantly, workers in AI-adopting firms have benefited through higher wages, both in aggregate and per employee. Whether these wage gains will persist in the long term, and whether they will be shared equitably across skill levels, remains an open question that merits continued monitoring. Uneven gains and the critical role of complementary investments Third, AI’s productivity benefits are far from evenly distributed. Breaking down our results by firm size reveals that medium and large companies experience substantially stronger productivity gains than their smaller counterparts (see Figure 2). This differential effect reflects the role of scale in absorbing AI integration costs and accessing complementary assets – data infrastructure, technical talent, and organisational capacity to redesign workflows. The finding raises concerns about widening productivity gaps between firms and regions, particularly given Europe’s industrial structure, which is dominated by small and medium-sized enterprises. Figure 2 Effects of AI adoption on labour productivity by company size Notes: The dependent variable is labour productivity, calculated as the log of turnover per employee, and is derived from EIBIS. AI adoption is measured using the AI implementation status derived from similar firms in the US. All regressions control for firm investment, profitability, financial leverage, total assets, age, and the interaction of country, sector and year fixed effects. Investment is expressed as the annual change in total fixed assets. Profitability is the ratio of earnings before interest and taxes (EBIT) to total assets. Financial leverage is the ratio of loans and long-term debt to total assets. All control variables come from Orbis and are lagged by 1 year. Error bars represent 90% confidence interval, based on standard errors clustered at country-sector-year level. Log linear approximation, 0.01 = 1% increase. Perhaps most importantly, our analysis reveals that AI adoption alone is insufficient. Firms must make complementary investments to unlock AI’s full potential. Our results show the striking heterogeneity in how different types of investments enhance AI’s productivity effects. An extra percentage point of investment in software and data infrastructure increases AI’s productivity effect by 2.4 percentage points. Investment in workforce training has an even larger multiplier effect: an additional percentage point spent on training amplifies AI’s productivity gains by 5.9 percentage points. These findings underscore a critical insight: the productivity dividends from AI depend not merely on acquiring the technology but on firms’ capacity to integrate it through investments in intangible assets and human capital. Implications for European policy These findings carry significant implications for policymakers. First, the benefits of AI adoption are mostly visible for medium and large firms. This means that Europe may work on policies that help smaller firms reach the critical scale necessary to benefit from AI. This requires well-functioning f