This is precisely the kind of structural question that will shape Europe's AI and technology future. It's also one of the themes at Human x AI Europe on May 19 in Vienna, where the intersection of governance capacity and technological sovereignty comes into sharp focus.
The Unanimity Trap
When someone argues that EU foreign policy is "broken," it helps to ask: broken compared to what? The Common Foreign and Security Policy (CFSP, the EU's framework for coordinating member states' external action) was never designed to function like a federal government's foreign ministry. It was designed to coordinate sovereign states that retain ultimate authority over their own security. The question is whether that design remains fit for purpose.
The evidence suggests growing strain. As recent analysis documents, single member states have repeatedly blocked collective action: Greece vetoed statements on China in 2017, Hungary blocked EU aid packages to Ukraine in 2022 and 2024, and Poland and Hungary jointly blocked migration statements in 2023. Each veto is legally permissible. Collectively, they raise a harder question: does the unanimity requirement protect unity, or does it weaponize disagreement?
The strongest argument for unanimity is that foreign policy touches the core of national sovereignty. States should not be bound to military commitments or diplomatic positions they oppose. The strongest argument against it is that the veto power has become asymmetric: states accused of undermining EU values can use procedural rules to prevent the Union from defending those values externally.
This is not a facts disagreement. Both sides acknowledge the same pattern of blocked decisions. It is a values disagreement about what the EU should be: a coordination mechanism among sovereign states, or an actor capable of independent strategic action.
The Security Imperative
The geopolitical context has shifted dramatically. As the German Institute for International and Security Affairs observed following the Munich Security Conference in February 2025, "it is evident that the United States has relinquished its role as Europe's security guarantor." The US reluctance to admit Ukraine into NATO and the absence of American security guarantees for European deployments have created what the institute calls "a significant challenge for European states."
This new reality has accelerated defence spending. According to analysis from the Swedish Institute for European Policy Studies, defence spending in EU member states has doubled since Russia's full-scale invasion of Ukraine in 2022. The European Commission's Readiness 2030 initiative aims to mobilize €800 billion in defence spending, with the SAFE (Security Action for Europe) instrument providing up to €150 billion in loans for joint procurement.
The money is flowing. The question is whether the institutional architecture can direct it effectively.
Three Positions, Three Logics
The debate over CFSP reform tends to cluster around three positions, each with its own internal logic:
Position One: Preserve Unanimity. Proponents argue that foreign policy legitimacy requires consent. Smaller member states, in particular, worry that qualified majority voting (QMV, where decisions pass with 55% of member states representing 65% of the population) would allow larger states to override their interests. As the Centre for European Reform notes, this concern intensifies with enlargement: "Moving away from unanimity would be one way of reducing the risk of stalemate, but this shift would be complex and faces much resistance, especially from smaller member-states concerned about losing influence."
Position Two: Expand QMV. Advocates contend that the current system is unsustainable. The EU already uses qualified majority voting for approximately 80% of legislation. Extending it to foreign policy would align decision-making with the Union's other policy areas. The existing QMV framework includes safeguards: a blocking minority requires at least four member states, preventing any single country from being overruled by a narrow coalition.
Position Three: Use Existing Flexibility. A middle path focuses on mechanisms already available under the treaties. Passerelle clauses allow shifts from unanimity to QMV without formal treaty change, though they require unanimous agreement to activate. Constructive abstention permits states to step aside without blocking collective action. Legal scholars argue that the EU could restore a "thin" Article 29 decision confined to essentials, leaving implementation details to qualified majority voting under Article 215.
What would have to be true for each position to be right? The unanimity defenders are right if the primary risk is large states dominating small ones. The QMV advocates are right if the primary risk is paralysis in the face of external threats. The flexibility proponents are right if the primary obstacle is political will rather than legal architecture.
The Technology Dimension
Foreign and security policy increasingly intersects with technology governance. The EU's AI Continent Action Plan, launched in April 2025, explicitly links AI development to "strengthening competitiveness and technological sovereignty." The Apply AI Strategy from October 2025 targets defence and space among its eleven priority sectors, with "European AI-enabled situational awareness systems through the European Defence Fund" and "dedicated computing infrastructure for training sovereign AI models by Q4 2026."
This creates a new category of foreign policy question. When AI systems become critical to military capability, decisions about AI governance become security decisions. When data flows cross borders, data protection becomes a matter of strategic autonomy. The Atlantic Council observes that "the links between AI and national security are a key issue for testing the strength of the transatlantic relationship."
The EU's regulatory approach to AI, embodied in the AI Act, was developed through ordinary legislative procedure with qualified majority voting. Its foreign policy implications, however, fall under CFSP and its unanimity requirements. This creates an awkward asymmetry: the Union can regulate AI internally by majority vote but may struggle to coordinate its external AI policy if any member state objects.
The Enlargement Complication
The debate gains urgency as enlargement returns to the agenda. Six countries are actively negotiating accession; several more are candidates or potential candidates. As the German Council on Foreign Relations reports, the Commission's proposals to fast-track Ukraine's membership have "prompted renewed scrutiny of enlargement policy by national governments."
An EU of 35+ members operating under unanimity in foreign policy would face arithmetic challenges. The more members, the more potential vetoes. The more diverse the membership, the harder to find consensus. Yet enlargement itself requires unanimous approval, creating a paradox: states that might benefit from blocking enlargement can do so, while states that might benefit from QMV cannot achieve it without unanimous consent.
Friends of Europe proposes a quid pro quo: the EU delivers simplification during pre-accession, while candidate countries accept transitional post-accession measures. This acknowledges that enlargement and institutional reform are linked, even if they proceed on separate tracks.
Where the Debate Actually Is
The European Parliament's January 2026 review of foreign, security, and defence policies captures the current state of play. MEPs warn that "an arc of instability has formed around Europe" and express "dismay at the current ambiguity of the EU-US defence and security relationship." They call for member states to be "ready to act autonomously on the basis of the European Union's mutual assistance clause."
The language is notable for what it does not say. It does not call for abolishing unanimity. It does not propose treaty change. It focuses on using existing mechanisms more effectively while acknowledging that the current architecture may be insufficient.
This is where the debate actually is: not a binary choice between unanimity and QMV, but a search for practical steps that can build capacity without triggering the political resistance that treaty change would provoke. The question is whether incremental adaptation can keep pace with accelerating external pressure.
The Question Worth Asking
The strongest version of the unanimity argument is that legitimacy matters more than speed. A foreign policy that lacks member state consent will lack implementation. The strongest version of the QMV argument is that paralysis is its own form of illegitimacy. A Union that cannot act loses credibility with allies and adversaries alike.
Both arguments contain truth. The question is which risk is more urgent in the current moment. And that question cannot be answered in the abstract; it depends on judgments about the threat environment, the reliability of allies, and the costs of inaction.
What the debate needs is not more position-taking but more precision. When someone says "the EU needs to strengthen its foreign policy," it helps to ask: strengthen against what threat? Through what mechanism? At what cost to other values? The answers will differ depending on whether the concern is Russian aggression, Chinese economic pressure, American unpredictability, or internal democratic backsliding.
The EU's foreign and security policy architecture was built for a different era. Whether it can be adapted for this one remains genuinely uncertain. That uncertainty is not a failure of analysis; it reflects the difficulty of the problem itself.
Frequently Asked Questions
Q: What is the EU's Common Foreign and Security Policy (CFSP)?
A: The CFSP is the EU's framework for coordinating member states' foreign policy positions and actions, established by the Maastricht Treaty in 1992 and consolidated by the Lisbon Treaty in 2009. It covers all areas of foreign policy and international relations, including the progressive definition of a common defence policy, and is led by the High Representative for Foreign Affairs and Security Policy.
Q: Why does EU foreign policy require unanimous voting?
A: Unanimity is required under Articles 24 and 31 of the Treaty on European Union because foreign policy touches core national sovereignty. The requirement ensures that no member state can be bound to military commitments or diplomatic positions it opposes. However, this also means any single state can veto collective action.
Q: What is the SAFE instrument and how much funding does it provide?
A: SAFE (Security Action for Europe), adopted by the Council in May 2025, provides up to €150 billion in competitively priced, long-maturity loans to member states for defence procurement. It is the first pillar of the Readiness 2030 plan, which aims to mobilize over €800 billion in total defence spending across the EU.
Q: Can the EU move to qualified majority voting without changing the treaties?
A: Yes, through passerelle clauses under Article 48(7) TEU, which allow shifts from unanimity to QMV without formal treaty amendment. However, activating these clauses requires unanimous agreement, creating a paradox: the states most likely to use vetoes must consent to removing their veto power.
Q: How does EU enlargement affect foreign policy decision-making?
A: Enlargement increases the number of potential vetoes under unanimity rules. With six countries actively negotiating accession and several more as candidates, an EU of 35+ members would face greater difficulty reaching consensus. This has intensified debates about whether voting reform should accompany or precede enlargement.
Q: What is the connection between AI policy and EU foreign and security policy?
A: AI increasingly affects military capability, intelligence, and strategic autonomy. The EU's Apply AI Strategy targets defence among its priority sectors, while the AI Act regulates AI systems internally. However, external AI policy coordination falls under CFSP unanimity requirements, creating asymmetry between internal regulation (majority voting) and external coordination (unanimity).