In August 2026, the European Commission’s Employment Expectations Indicator reached its highest level since January, offering a positive signal for future office demand. But what is driving the improvement: firmer economic growth prospects, or the expansion of AI-related activity across Europe’s service sector?
Savills analysis compares 2025 business AI adoption rates with the EU’s August 2026 service-sector employment expectations indicator across 12 European countries. To separate the relationship with AI adoption from the effect of the economic cycle, the model controls for Oxford Economics’ average GDP growth forecasts for 2026–27.
Holding forecast GDP growth constant, each one-percentage-point increase in business AI adoption is associated with a 0.20-point increase in services employment expectations. Across the 12 countries analysed, AI adoption and employment expectations have a correlation coefficient of +0.79. Together, these results suggest that AI adoption may be an important feature of the service-sector recovery.
Belgium, Denmark and Sweden stand out as markets where relatively high AI adoption is accompanied by stronger employment sentiment. By contrast, several lower-adoption markets report weaker employment expectations, even after accounting for GDP growth. This does not prove causality, but it challenges the simplistic idea that AI adoption automatically translates into weaker office-based employment.
For office markets, AI is likely to reshape demand more than reduce it. In the near term, higher-adoption markets may also be those where service-sector firms are investing, expanding and recruiting. Demand is therefore likely to concentrate in cities with deep technology ecosystems, skilled labour pools, access to capital and the infrastructure needed by digitally intensive businesses. Centrally located, well-connected areas are therefore becoming all the more sought-after by occupiers.
There are, of course, important caveats. Political uncertainty, fiscal policy and sector composition can all influence hiring expectations. For example, the run-up to France’s 2027 presidential election and uncertainty around the UK’s Autumn Budget may weigh on sentiment, while stronger tourism activity in parts of southern Europe could support employment expectations above modelled levels. AI adoption may also be acting as a proxy for wider structural advantages, including digital maturity, skills availability, firm size and sector mix. The relationship should therefore be treated as evidence of association rather than proof that AI adoption is directly causing stronger hiring sentiment.
The bottom line is that the early evidence does not conclude that “AI equals fewer offices”. Instead, higher AI adoption is currently associated with stronger service employment expectations, even after controlling for forecast GDP growth. For offices, the more likely near-term effect is a shift in where demand goes, what type of space occupiers want, and which buildings are best positioned to capture it.
Contact Mike Barnes
AI and European Offices: Separating Hype from Hiring Reality
.jpg){kind=logo}
