Today, 25.05.2026
Good morning, Human. The European Commission has spent the past year producing AI strategies, action plans, and competitiveness compasses. Now comes the part that actually matters: figuring out whether any of it works. A Brussels think tank is about to find out.
In Brief
The CEPS Task Force on the Apply AI Strategy launches next month, bringing together industry, policymakers, and researchers to stress-test the EU's ambitious AI deployment plans across healthcare, automotive, and government sectors. The initiative matters because the Apply AI Strategy arrived with no dedicated budget and unresolved questions about infrastructure, data sharing, and whether flagship projects will serve real industrial needs. For European decision-makers, this is where the gap between policy documents and production reality either closes or becomes permanent.
If this is the conversation that matters to you, it's happening in person soon. Human x AI Europe convenes in Vienna on May 19, bringing together the people who actually have to make these strategies work.
The Lead: When Strategy Meets Reality
The Centre for European Policy Studies (CEPS) is launching a dedicated Task Force to do something the European Commission's Apply AI Strategy couldn't: answer the hard questions. Running from June 2026 through February 2027, the initiative will produce a final report, three sector-specific policy briefs, and a stakeholder network connecting European Digital Innovation Hubs (EDIHs), industry players, and policymakers.
The Task Force operates across three sector tracks: Healthcare & Pharma, Automotive & Mobility, and Government & Public Sector. Each track will apply a consistent analytical framework across five issue areas: key use cases, infrastructure requirements, data governance, flagship applications, and sovereignty and implementation roadmaps. The structure is deliberate. CEPS Director of Research Andrea Renda has been direct about the challenge: the Apply AI Strategy carries a heavy implementation burden, no dedicated budget, and unresolved questions about whether existing initiatives will actually serve industrial needs.
Here's the mechanism hiding under the headline. The Apply AI Strategy, launched in October 2025, sets out to scale AI deployment across eleven strategic sectors. It adopts an ecosystem approach, connecting use cases with infrastructure, data governance, and skills needs. It avoids one-size-fits-all solutions in favor of tailored, sector-specific flagships. It keeps a firm link with the broader EU legislative agenda, including the AI Act and the Cloud and AI Development Act. On paper, this sounds like exactly what Europe needs.
The problem is execution. The EU's data strategy has struggled to produce results so far. From the European Health Data Space to Catena-X in the automotive sector, industry hasn't found a good way to share data in a way that supports primary and secondary data uses. The AI Factories and Gigafactories initiative, while ambitious, faces questions about whether the infrastructure is being placed optimally. A CEPS analysis from November 2025 found that factories are being built mostly outside of AI "hubs of excellence" and are not leveraging the most favorable energy conditions in Europe.
The Task Force will operate under Chatham House Rules, which means participants can speak freely without attribution. This matters. The gap between what companies say publicly about EU AI policy and what they say privately is substantial. CEPS is betting that structured, confidential engagement can surface the tensions that polite policy documents paper over.
The Infrastructure Question
The Apply AI Strategy's infrastructure ambitions are substantial. The Commission has committed to building four to five AI Gigafactories within the EU, with an investment of €20 billion through the InvestAI Facility. Currently, 19 AI Factories and 13 Antennas are operational, with at least nine new AI-optimized supercomputers to be procured and deployed across the EU.
But industry groups have raised pointed questions about the business case. According to Science|Business reporting, many companies don't see a clear path to profitability for gigafactory operations. The question isn't whether Europe needs more AI hardware. The question is whether anyone will come to use it, and whether the consortia operating these facilities can sustain them financially.
The Future Society's March 2026 analysis of Europe's AI policy portfolio identified a structural vulnerability: computing infrastructure is concentrated within a limited number of flagship projects. With just five flagship initiatives accounting for over €30 billion in targeted compute investment, any failure to deliver on time or at scale leaves Europe without a distributed fallback. The schedules for infrastructure delivery and the launch of research programs dependent on that infrastructure are not explicitly aligned.
The Security Situation
While Europe debates AI deployment strategy, the security landscape is shifting beneath everyone's feet. Google's Threat Intelligence Group published a sobering report on May 11 documenting what it believes is the first case of a threat actor using AI to develop a working zero-day exploit intended for mass exploitation. Google says it discovered the operation before the attacker could deploy it broadly, giving the software developer time to fix the issue first.
The report signals a maturing transition from nascent AI-enabled operations to industrial-scale application of generative models within adversarial workflows. Threat actors associated with China and North Korea have demonstrated significant interest in capitalizing on AI for vulnerability discovery. AI-driven coding has accelerated the development of polymorphic malware by adversaries. Autonomous malware operations, where models interpret system states to dynamically generate commands, are no longer theoretical.
For European organizations deploying AI at scale, this creates a dual challenge. The Apply AI Strategy encourages an "AI first" policy where AI is considered as a potential solution whenever organizations make strategic decisions. But every AI deployment expands the attack surface. According to Darktrace's State of AI Cybersecurity 2026 survey, 92% of security leaders are concerned about the use of AI agents across the workforce and their impact on security. The rapid expansion of generative AI across the enterprise is outpacing the security frameworks designed to govern it.
The Healthcare Track
Healthcare is where the Apply AI Strategy's ambitions meet the most complex regulatory and operational environment. The Commission has announced plans to establish European AI-powered advanced screening centers by Q2 2027, building on the European Health Data Space and digital infrastructures for cancer imaging and genomic data.
A new Digital Europe Programme call (DIGITAL-2026-AI-PILOTING-10-SCREENING) with a €9 million budget aims to pilot cloud-based AI systems for medical image screening, focusing on early detection of cancer and cardiovascular diseases. The deadline for submissions is October 1, 2026. Two projects will be funded, each receiving up to €4.5 million at a 50% funding rate.
The call's requirements reveal the gap between AI research and clinical deployment. Successful projects must develop secure, privacy-preserving cloud environments fully interoperable with existing health IT infrastructure. They must demonstrate large-scale validation across vast numbers of patient cases in real-world healthcare settings. They must ensure human-in-the-loop validation by qualified medical professionals. And they must align with the EU AI Act, cybersecurity frameworks, and data protection rules while preparing documentation for Health Technology Assessment and regulatory approval.
This is where the CEPS Task Force's healthcare track becomes critical. The technical capability to deploy AI in medical imaging exists. The question is whether Europe's fragmented health data infrastructure, varying national regulations, and institutional resistance can be overcome at scale.
The Week Ahead
May 26-27: EuroDIG, the Pan-European Dialogue on Internet Governance, convenes in Brussels. AI governance will feature prominently in discussions about Europe's digital future.
May 27: The Academy of European Law hosts a webinar on the EU Savings and Investments Union, with implications for how Europe finances its AI infrastructure ambitions.
May 27-29: The 2nd European Forum on Digital Citizenship Education takes place in Strasbourg, addressing how AI literacy fits into broader digital competency frameworks.
Ongoing: The Digital Omnibus on AI amendments, provisionally agreed on May 7, continue their path through EU institutions. The package includes a 16-month deferral of high-risk AI system obligations under Annex III, pushing the deadline from August 2026 to December 2027.
The Thought That Lingers
The Apply AI Strategy represents Europe's attempt to move from regulating AI to deploying it. The CEPS Task Force represents something more modest but perhaps more important: an honest assessment of whether the strategy's assumptions hold up when tested against industrial reality. The Commission has produced the vision. Now a room full of practitioners, operating under Chatham House Rules, will determine whether that vision can survive contact with the world as it actually is. The answer will shape whether Europe's AI ambitions become operational capability or remain, as one CEPS analysis put it, "cathedrals in the desert."
Frequently Asked Questions
What is the CEPS Task Force on the Apply AI Strategy?
The CEPS Task Force is a nine-month initiative running from June 2026 through February 2027 that brings together industry, policymakers, and researchers to evaluate the EU's Apply AI Strategy. It operates across three sector tracks (Healthcare & Pharma, Automotive & Mobility, and Government & Public Sector) and will produce a final report, three sector-specific policy briefs, and a stakeholder network connecting European Digital Innovation Hubs with industry players and policymakers.
Why does the Apply AI Strategy need evaluation?
The Apply AI Strategy launched in October 2025 with ambitious goals to scale AI deployment across eleven strategic sectors, but it came with no dedicated budget and unresolved questions about infrastructure placement, data sharing mechanisms, and whether flagship projects will actually serve industrial needs. The CEPS Task Force aims to stress-test these assumptions against real-world implementation challenges.
What are the main challenges facing EU AI infrastructure development?
The EU faces several infrastructure challenges: AI Factories and Gigafactories are being built mostly outside of AI "hubs of excellence" and aren't leveraging optimal energy conditions; industry groups question the business case for gigafactory operations; and computing infrastructure is concentrated in just five flagship initiatives accounting for over €30 billion, creating vulnerability if any fail to deliver on time or at scale.
How is AI changing the cybersecurity landscape?
Google's Threat Intelligence Group documented the first case of a threat actor using AI to develop a working zero-day exploit for mass exploitation. AI-driven coding has accelerated polymorphic malware development, and autonomous malware operations are no longer theoretical. According to Darktrace's 2026 survey, 92% of security leaders are concerned about AI agents' impact on security as generative AI expansion outpaces security frameworks.
What are the specific challenges for AI deployment in European healthcare?
Healthcare AI deployment faces complex regulatory and operational challenges including fragmented health data infrastructure, varying national regulations, and institutional resistance. While the Commission plans AI-powered screening centers by Q2 2027 and has allocated €9 million for medical imaging AI pilots, successful projects must navigate EU AI Act compliance, cybersecurity frameworks, data protection rules, and prepare for Health Technology Assessment approval.
Human×AI Daily Brief is compiled from CEPS, the European Commission's Digital Strategy portal, Google Threat Intelligence Group, Darktrace, Science|Business, The Future Society, and the Council of Europe. This is meant to be useful, not comprehensive.