Europe’s scale-ups are raising and hiring to build AI, but who are they hiring to govern it?

It’s no surprise that European companies are racing to build AI and hiring accordingly, but according to new data, the people needed to govern those systems are not being hired at the same pace.

New research from compliance consultancy Axipro found that companies across eight EU countries advertised 6.7 AI builder roles for every AI governance role, highlighting a widening gap between Europe’s technological ambitions and its capacity to manage the risks that follow.

Axipro analysed 3,519 English-language LinkedIn job advertisements published between 1 June and 1 July 2026 across Belgium, France, Germany, Ireland, Italy, the Netherlands, Spain and Sweden.

The sample included 3,004 roles focused on building AI systems and 446 governance positions. More strikingly, 71.5% of the governance job descriptions did not explicitly mention the EU AI Act , despite it being the bloc’s central framework for regulating AI.

While the figures do not prove that European startups are disregarding regulation (Axipro’s research measures the flow of new hiring rather than the size or quality of existing teams), the disparity offers a useful signal: Europe’s investment in AI products, infrastructure and technical talent appears to be moving faster than its investment in the operational capacity required to govern them.

“ Job postings are the most honest signal a company sends. Strategy decks say what leadership would like to do. Hiring says what they’re actually doing ,” says Ali Hayat , founder and CEO of Axipro.

Uneven across Europe

The imbalance was visible in every country studied, although its scale varied considerably.

Sweden recorded the widest gap, with 16 AI builder positions advertised for every governance role. France followed at 11.4 to one, while Belgium stood at 7.9 to one and the Netherlands at 7.2 to one. Ireland had the most balanced market, but even there companies advertised 3.5 builder roles for every governance position.

Explicit references to the EU AI Act were also inconsistent. Italy had the highest proportion of governance vacancies naming the legislation, at 45%, followed by Belgium at 39.5% and France at 38.5%. Ireland, despite having the strongest overall balance between technical and governance hiring, had the lowest explicit mention rate, at 14.6%.

Some governance job descriptions referred instead to related frameworks and standards such as GDPR, DORA and ISO 27001. These skills can contribute to responsible AI operations, but their inclusion does not necessarily demonstrate that a company has mapped its systems against the AI Act’s specific requirements.

The scaleup squeeze

The consequences are likely to be felt most sharply by scale-ups and mid-market businesses. These companies are often sufficiently established to sell to large organisations, operate in regulated sectors or deploy AI across important workflows.

At the same time, they may lack the budgets and organisational depth needed to maintain standalone legal, risk, security and AI governance teams.

The data identifies businesses with between 30 and 300 employees as particularly exposed. Mid-market companies represented 34% of the governance hiring found in the study, compared with 48% for large enterprises and 18% for small companies.

Yet regulatory and commercial expectations do not necessarily shrink in proportion to a company’s workforce.

“ Large enterprises have compliance departments. Small companies mostly fall outside scope. The exposed group is the 30 to 300-person firm: regulated like the big players, staffed like the small ones ,” adds Ali .

Capital is flowing fast

This hiring gap is developing against a backdrop of strong 2026 capital flows into European AI companies.

EU-Startups’ 2026 coverage includes major disclosed rounds for AI-focused businesses such as Nscale , which secured €1.7 billion, Skello with €200 million, Viktor with €64.7 million, Conduct with €51 million and SPREAD with €25 million.

Other funded AI companies active during the year include Orbio, Zepo Intelligence, LawX, Redpine, Delfos Energy, Mafer AI, Kopa.ai, Vox Talk AI and Vectrix – to name a few.

AI is also increasingly embedded as an enabling capability in businesses working across fleet management, climate technology, payroll, aviation maintenance, energy and waste sorting.

Governance as a product feature

Investment in businesses specifically addressing governance and compliance is visible too, although it remains a smaller part of the funding picture. NeuralTrust raised €17.2 million to secure and govern enterprise AI agents, while Hybridity secured €2 million for its automated compliance technology. Cleo Labs raised €1.5 million to support international product compliance, and Cortea secured €12 million to develop AI quality controls for audit firms.

The emergence of these companies reflects a broader change in how governance is being viewed. Compliance is not only a legal exercise intended to prevent enforcement action. For AI vendors, it is increasingly tied to sales, procurement and customer confidence.

Larger corporate buyers may expect suppliers to explain how models are tested, how data is handled, who remains accountable for automated decisions and which records can be produced when problems arise.

A scaleup unable to answer those questions may find that its sales process slows, even in the absence of direct regulatory action. Governance capacity can consequently influence whether an AI product is trusted, approved and purchased.

“ Everyone’s pricing this as a fines question, and I think that’s the mistake. Yes, enforcement will be selective – but selective enforcement still needs examples, and nobody controls whether they’re chosen. The bigger shift is commercial. Compliance is becoming a product feature ,” says Ali .

Closing the gap

Europe’s challenge is therefore not simply to produce more governance specialists for every AI engineer. Technical and governance work do not need to grow at identical rates, and companies will require different structures depending on their size, sector and use of AI.

The more important issue is whether governance capabilities are growing quickly enough to support the systems already being developed and deployed.

Europe has demonstrated that it can fund and build ambitious AI companies. Its next test is whether those companies, particularly those moving from startup to scaleup, can develop the governance foundations needed to grow without losing the trust of their users.