In early June 2026, the Committee of Permanent Representatives (COREPER) authorized the European Commission to formalize the EU’s participation in Pax Silica, the US-led semiconductor supply-chain initiative. By June 26, the European Union was a formal signatory.

The process was orderly. It was also, by any meaningful democratic standard, invisible. No European Parliament debate preceded the authorization. No impact assessment was published. No connection was drawn to a parallel development: the entry into force of the GENIUS Act, the US stablecoin law whose implementing rules were due July 18 and whose extraterritorial provisions reach into European financial infrastructure. The EU bound itself to one American architecture while remaining structurally exposed to another. Both decisions unfolded within the same three-week window, and neither received public scrutiny proportionate to its implications.

What the EU Signed

Pax Silica was launched on December 12, 2025, with seven founding signatories. By the second summit in Washington on June 25–26, it had grown to 24, with the EU joining as an institution alongside Germany, the Netherlands, and seven others. The initiative establishes a US-led framework for semiconductor supply-chain governance, workforce development (the Foundry School, in partnership with Stanford), and a proposed $250 million fund for critical minerals extraction.

The accession was not naive. COREPER deliberations involved negotiation, and several Member States raised concerns. But the process by which the EU joined exposes a structural problem in how the EU governs its integration into US-designed frameworks.

COREPER’s own preparatory documents classify Pax Silica as a “nonbinding instrument.” Under established CJEU case law and Council practice, nonbinding political declarations fall outside Article 218 TFEU – the Treaty provision that governs the negotiation and conclusion of international agreements between the EU and third countries or international organisations. Under Article 218, the Council authorises the opening of negotiations (paragraph 2), the Commission conducts them, and the Council concludes the agreement – but only after obtaining the consent of the European Parliament where the agreement covers fields subject to the ordinary legislative procedure, or after consulting it in all other cases (paragraph 6). The Court of Justice has reinforced this architecture through a principle of symmetry: Parliament’s external treaty powers must mirror its internal legislative powers, ensuring that it cannot be sidelined on international commitments in fields where it would be a full co-legislator domestically (Case C-658/11, Parliament v Council). Since Pax Silica is classified as nonbinding, Article 218 does not apply. Accordingly, no European Parliament consent is required. The procedure was formally correct.

But the distance between formal classification and operational reality is striking. Pax Silica has a dedicated secretariat function, a proposed $250 million funding mechanism, a workforce development program co-designed with Stanford, and a membership structure that distinguishes signatories from non-signatories. It is, in practice, an institutional architecture with expanding governance features. What is more, the EU routinely participates in similar nonbinding frameworks – the Minerals Security Partnership, Global Gateway memoranda of understanding – without triggering Article 218. The pattern is consistent: structural commitments are designed to remain formally nonbinding precisely because that classification avoids parliamentary scrutiny.

The risk is not merely procedural. Once inside a framework of this kind, exit becomes costly through mechanisms that operate independently of legal obligation. Financial contributions to the $250 million fund create budgetary expectations that are difficult to reverse without diplomatic consequences. Workforce integration through the Foundry School embeds European participants in US-designed training pipelines whose standards, once adopted, function as de facto regulatory requirements for supply-chain participation. The membership structure itself generates a dynamic of inclusion and exclusion: non-signatories risk being treated as less reliable partners in semiconductor supply-chain allocation. None of these effects requires a binding treaty. They arise from institutional participation itself – and they accumulate precisely because no single commitment, taken in isolation, crosses the threshold of legal bindingness.

This matters because of the symmetry principle. Semiconductor industrial policy, supply-chain governance, and critical minerals extraction all fall within fields where the ordinary legislative procedure applies internally. The European Chips Act (Regulation 2023/1781) was adopted through precisely that procedure, with full European Parliament co-decision. If the EU were negotiating a binding international agreement on the same subject matter, Article 218(6)(a) would require Parliament’s consent. The nonbinding classification is the mechanism that prevents that requirement from triggering – and it is the only mechanism.

The question, then, is not whether Article 218 applies. It does not, as the instrument is currently classified. The question is whether a classification system that permits the EU to join governance architectures with dedicated funding, workforce integration, and institutional infrastructure – all without European Parliament involvement – is democratically adequate for commitments of this structural significance.

What the EU Did Not Address

Three weeks before the EU signed Pax Silica, the GENIUS Act’s regulatory implementation entered its final phase. The GENIUS Act (Guiding and Establishing National Innovation for US Stablecoins), signed into law on July 18, 2025 with the broadest bipartisan majority on any financial legislation in the 119th Congress, is the first comprehensive US federal law governing stablecoins – digital tokens designed to maintain a stable value, typically pegged one-to-one to the US dollar and backed by reserves of cash or Treasury securities. Stablecoins are not a niche product: the global market exceeds $300 billion, with dollar-denominated instruments representing approximately 95% of the total. They function, in practice, as privately issued digital dollars that can be transferred globally, instantly, and – until the GENIUS Act – largely outside the perimeter of any single regulatory authority.

The GENIUS Act changes this by establishing US Treasury and OCC (Office of the Comptroller of the Currency) oversight over all stablecoin issuers – including those domiciled outside the United States – that serve US persons or operate on US-regulated platforms. Crucially, the Act requires every issuer to maintain the technical capability to comply with lawful US orders to freeze or seize assets. A “lawful order” under the Act is defined as any final and valid order issued by a US court or authorized federal agency requiring the seizure, freezing, or prevention of transfer of payment stablecoins. Such orders are subject to judicial or administrative review only after execution, not before. Non-compliant foreign issuers can be designated by Treasury and prohibited from offering their instruments to US persons on US-regulated platforms – which, given the dominance of US platforms in global stablecoin trading, would in practice severely constrain their liquidity worldwide. Seven US federal agencies were required to publish implementing rules by July 18, 2026. None met the statutory deadline. Full operational effect is scheduled for January 18, 2027, but the regulatory framework that should govern it remains incomplete.

This creates a specific and measurable exposure for European financial institutions. Dollar-denominated stablecoins used in European markets – whether held directly by banks, used as settlement instruments, or intermediated through EU-licensed platforms – operate under US jurisdictional authority. A concrete example illustrates the mechanism: if the US Office of Foreign Assets Control were to blacklist USDT funds traceable to a European bank, the issuer (Tether) would be required to freeze those funds to remain compliant with the GENIUS Act. The European institution would lose access to those assets immediately, with legal recourse available only after the fact. This is not a theoretical capability: Tether itself froze $514 million in USDT during 2026 in response to law enforcement requests.

The EU’s own regulatory framework does not address this exposure. MiCA (Markets in Crypto-Assets Regulation, EU 2023/1114), which entered into force in 2024, is Europe’s comprehensive framework for crypto-asset markets. It imposes reserve requirements, authorization procedures, and oversight by the European Banking Authority (EBA) and the European Securities and Markets Authority (ESMA) on crypto-asset issuers established within the European Union. MiCA is a sophisticated instrument, but it governs EU-domiciled issuers only. It does not govern non-EU issuers such as Tether, the largest stablecoin by market capitalisation (approximately $190 billion in circulation, roughly 59% of the global stablecoin market), which is domiciled in El Salvador. MiCA contains no mechanism for EU institutions or individuals to contest or delay compliance with a US lawful order affecting a non-EU stablecoin issuer. No bilateral treaty, WTO framework, or equivalent instrument currently fills this gap.

The scale of the exposure is itself unknown. The ECB and EBA do not publish aggregate data on how much EU-supervised institutions hold in stablecoins issued by non-EU entities. Without that figure, the gap cannot be scoped or priced. On May 9, 2026, ECB President Lagarde publicly warned that large stablecoins including Tether and USDC pose financial stability risks and could transmit stress to underlying asset markets. That warning was not accompanied by any regulatory instrument capable of addressing the specific extraterritorial mechanism created by the GENIUS Act.

The ECB’s near-term technical response is Pontes, a settlement solution based on distributed ledger technology (DLT) – the shared digital record-keeping systems, of which blockchain is the best-known example, that allow multiple participants to verify and record transactions without a central intermediary. Pontes is designed to connect market DLT platforms with the Eurosystem’s existing TARGET payment services, enabling wholesale financial transactions to settle in central bank money rather than in private stablecoins. Its initial launch is not due until Q3 2026. The broader Appia initiative – the ECB’s longer-term project to build an integrated tokenised wholesale financial ecosystem for the eurozone – will not deliver its ecosystem blueprint until 2028. The transitional grandfathering period under MiCA expired on July 1, 2026, but MiCA does not address the extraterritorial reach of the GENIUS Act’s freeze authority.

The result is a structural gap: European financial institutions using dollar-denominated stablecoins are subject to US enforcement authority with no European fallback infrastructure in place and no EU-level legal framework addressing the interaction between MiCA and the GENIUS Act.

This gap is not speculative. Eleven companies filed for or received OCC national trust bank charters in 83 days between December 2025 and March 2026, including firms with substantial European operations such as Circle, Ripple, and Crypto.com. The infrastructure for the GENIUS Act’s operational phase is being built now.

The Democratic Governance Question

The concern is not that the EU made the wrong decision on Pax Silica, or that the GENIUS Act exposure is necessarily catastrophic. The concern is procedural: the EU is integrating into US-designed governance architectures across multiple domains – semiconductor supply chains, digital financial infrastructure, AI export controls – through executive channels that bypass the ordinary mechanisms of democratic scrutiny.

Each decision is handled in its own silo: Pax Silica through COREPER and the Council, MiCA through the ordinary legislative procedure (completed in 2023), GENIUS Act exposure through no EU-level process at all. No institution is assessing the cumulative effect. The European Parliament has not debated the relationship between these frameworks. No Commission communication has examined how Pax Silica membership, GENIUS Act exposure, and the EU’s own MiCA framework interact as a system.

The silo structure is not accidental. Each Council configuration – competitiveness, economic and financial affairs, foreign affairs – sees only the commitments that fall within its sectoral mandate. COREPER coordinates horizontally, but as the Pax Silica accession demonstrates, coordination produces process management, not strategic assessment. When the Trade and Technology Council, the Competitiveness Council, and ECOFIN each handle their respective piece of the US architecture, no institutional actor is positioned to ask whether the aggregate effect amounts to a structural dependency that the EU has not debated, voted on, or even named.

This is a governance gap, not a policy gap. The substantive merits of each decision may be defensible. But the absence of an integrated assessment means that the EU is making structural commitments to US-designed architectures without a democratic process that matches the scale of the commitment.

What Should Change

Three measures would address the procedural deficit without requiring the EU to reverse any substantive decision.

First, the Commission should publish an assessment of the Pax Silica Declaration’s evolving institutional features – its funding mechanism, workforce programs, and secretariat functions – and evaluate whether these features, individually or cumulatively, warrant reclassification of the instrument under Article 218 TFEU or, at minimum, a voluntary referral to the European Parliament for debate. The principle of symmetry established by the Court of Justice provides the doctrinal basis for such an evaluation: where the subject matter of an international commitment falls within a field governed internally by the ordinary legislative procedure, Parliament’s role in the external dimension should not be circumvented through the formal classification of the instrument as nonbinding.

Second, the European Parliament should request a Commission assessment of the interaction between MiCA and the GENIUS Act, specifically addressing the extraterritorial reach of US freeze authority over dollar-denominated stablecoins used by European financial institutions and the adequacy of the Pontes/Appia timeline as a response.

Third, the Council should establish a horizontal review mechanism for EU participation in US-led governance frameworks, ensuring that commitments made through COREPER in different sectoral configurations are assessed for their cumulative structural implications.

The EU’s participation in international semiconductor cooperation is a legitimate policy choice. Its exposure to the GENIUS Act’s extraterritorial provisions is a manageable risk – if managed. But managing these commitments requires a democratic process capable of seeing them as connected. At present, no such process exists.