EU special envoy for industrial AI Jim Hagemann Snabe told the 27 European commissioners at a private seminar on 4 September that Europe should focus on using AI to lift slow productivity rather than racing to build the best models. The account, reported for Politico by Zoya Sheftalovich and attributed to two Commission officials, named health, farming, transport, defence and factories as priority areas for deployment. The signal matters for business because adoption support, not model building, looks set to anchor coming EU industrial measures.
What Snabe told the 27 commissioners
A week before the 4 September seminar, a Commission AI expert gave a similar briefing to the commissioners' heads of cabinet, according to Politico. The line in slides seen by Politico was that the US and China may lead on models and computing power, while Europe can still gain by putting AI to use. Commission spokesperson Thomas Regnier described the discussion as an exchange on how Europe can get firms to adopt AI faster. Taken together, the two briefings show a coordinated message delivered first at staff level and then directly to the 27 commissioners.
The slides pointed to sector-specific enablers rather than general funding promises. Examples included easier testing rules for self-driving cars, secure data spaces for personalised medicine and a digital adviser for farmers. Each case links a regulatory or data condition to deployment: testing rules for transport, governed data sharing for health, and advisory tools for farming. Commission President Ursula von der Leyen echoed the approach in her State of the Union speech on 16 September, saying Europe does not need to build frontier AI to get the most value from it.
Snabe, who chairs the supervisory board of Siemens, was appointed in June to advise von der Leyen and tech chief Henna Virkkunen, an appointment that drew conflict-of-interest criticism at the time. The adoption-first position is not uncontested inside Europe. German MEP Axel Voss said the Commission was too slow to enforce its AI law and pointed to recent hacks by AI agents, Politico reported. Earlier this month, former commissioner Margrethe Vestager and economist Philippe Aghion called for Europe to hold 15% of global compute by 2030, a capacity goal that sits apart from the use-focused line.
What this means for companies using AI
For companies operating in the EU, an adoption-centered agenda would shift attention from which model to buy to how quickly a tool can be deployed inside existing operations. In transport, simpler testing regimes could shorten pilots for self-driving systems; in health, secure data spaces could define how personalised medicine projects access and share data; in farming, a digital adviser points to ready-made assistance rather than in-house development. Larger firms with compliance and data teams could move first under such frameworks, while smaller firms would feel the effect mainly if shared spaces, advisers and testing procedures lower integration costs.
The same outline leaves several conditions that buyers and managers need to verify. Enforcement of the EU AI law remains disputed, with Voss warning about slow implementation at a time of hacks involving AI agents, which raises due-diligence requirements for agent deployments. The Siemens link that prompted criticism around the appointment also means firms should track how sector advice translates into neutral rules. The news by itself does not change testing regimes, data access or funding; Snabe's team will study sector use until December and draft policies near the end of the year, with a final report and strategy expected in early 2027.
The marker to watch is that timetable: sector studies through December, draft policies near year-end, then a final report and strategy in early 2027. If drafts convert the three slide examples into concrete testing, data-sharing and advisory mechanisms, the adoption line will have held. If the debate swings toward the 15% compute target, capacity policy will compete with it for budget and attention.
