On 6 October 2026, the Commission presented its Communication “Preparing for a wider Union”, which sets out concrete measures to prepare the Union for enlargement. At the centre are so-called institutional safeguard clauses. These resemble an Article 7 TEU procedure for the late arrivals – but at a discount, featuring a much broader scope and much lower thresholds. I argue that this is not the way. Let me be clear: the Commission’s concern, protecting the Union’s values and guaranteeing its agency after accession, deserves support. My quarrel is with the means. While it is doubtful whether the proposed clauses still qualify as an “adjustment” of the current Treaties under Article 49(2) TEU, they strike at core principles of the Union’s constitutional framework: the equality of Member States and, behind them, the equality of Union citizens. Instead of creating an unequal membership, the Commission should find solutions that apply to all Member States, whether under the existing Treaty framework, through accession treaties, or – if necessary – through Treaty amendment.

Dilemma: Enlargement Yes, Reform No?!

The political momentum for enlargement is strong: Ukraine, Albania, Montenegro, and Moldova are set to join the Union in the coming years. The Commission President speaks of a “geopolitical imperative”. At the same time, the Union is far from ready for enlargement. It struggles with deadlocks in many areas, especially when it comes to the protection of values and its capacity to act in sensitive fields such as the CFSP and the budget. Orbán’s spectre still haunts Europe … Further accessions will make the Union not only larger but also more heterogeneous. Decision-making structures that struggle to work in an EU of 27 are even less likely to work in an EU of 30+.

Two ways lead out of this dilemma. One is to differentiate membership, the other to change the Union’s framework. The first has been suggested in the form of associate membership, which Ukraine dismissed – understandably – as unfair: “We are defending ​Europe – fully, not partially, and not with half-measures”, Zelenskyy said. The second is politically even harder. The Member States show little appetite for reform. In 2023, the European Parliament proposed a large-scale overhaul of the Treaties, so far to no avail. The catalytic force that enlargement once exerted on the Union’s constitutional development through Treaty change seems, at least for the moment, to have disappeared.

Attentive to the political mood, the Commission stressed in 2024 that the institutional framework “can be swiftly improved by using to the full the potential of the current Treaties”. Grudgingly, the European Parliament gave its blessing to this approach à droit constant, acknowledging that institutional pre-enlargement reforms can be achieved “by activating the flexibilities offered by the current Treaties”.

The 2026 Communication reiterates this approach – “[t]he Lisbon Treaty was designed … to cater for further enlargements”, we are told (p 3) – and sketches “targeted adaptations within the existing Treaty framework and accession treaties” (p 2). While the suggested adaptations of the framework are hardly surprising (passerelle clauses, institutional practices, enhanced cooperation …), the heavy lifting seems to be left to the accession treaties.

Proposal: Institutional Safeguard Clauses – A Parallel Article 7 TEU Procedure for New Member States?

What does the Commission’s proposal entail? New accession treaties would address breaches affecting the Union’s “institutional functioning” or “membership obligations” through an institutional safeguard clause (pp 21 f). Substantively, it would cover breaches by a new Member State not only of the values in Article 2 TEU, but also of the principle of sincere cooperation. Procedurally, the Commission would be empowered to propose measures that “would be deemed adopted unless the Council rejected them by qualified majority within a specified period”. Possible responses could even extend to the suspension of voting rights in the Council.

This proposal seems to kill two birds with one stone, addressing both the Union’s protection of values and its capacity to act. By extending the clause to breaches of the malleable principle of sincere cooperation, the Commission may also capture obstruction in the Council. One immediately thinks of Orbán’s hostage-taking in the CFSP. In such situations, the Commission could allege a breach of sincere cooperation and propose suspending the obstructing state’s voting rights. And the procedural arrangement is favourable, to say the least: Member States would need a qualified majority to stop the Commission.

Ultimately, this amounts to a parallel Article 7 TEU procedure – but with a much broader substantive scope and much lower procedural thresholds. Anticipating objections, the Commission seeks to reassure us. This “safeguard would complement, not replace, Article 7 TEU … Article 7 would retain its central role in protecting the Union’s values”. But why, one wonders, would we need Article 7 TEU if we had this new tool at our disposal? Further, the Commission stresses the temporary nature of this mechanism, which would apply for 15 years after accession. “Only?” one is tempted to exclaim. Finally, any measure would be subject to judicial scrutiny. Again, one wonders: full review, or merely procedural review along the lines of Article 269 TFEU?

Legality: Treaty Adjustments Through Accession Treaties

The obvious question is: is this legal? The Member States can do quite a lot in accession treaties. Article 49(2) TEU provides a route to change primary law, allowing “adjustments to the Treaties on which the Union is founded”. Accordingly, accession treaties have the rank of primary law (see eg Austria v Council, C-445/00, para 62).

What do such “adjustments” entail? First and foremost, they include prosaic changes to the territorial scope of EU law (Article 52 TEU) or to the language versions of the Treaties (Article 55 TEU). Second, they cover transitional rules that slow the pace of accession in certain areas. Examples include restrictions on free movement in relation to Bulgaria and Romania (see eg Article 23 of the Bulgarian and Romanian Accession Act, and Chapter 1 of Annexes VI and VII). And third, accession acts have contained safeguard clauses, both for the internal market and the AFSJ (see eg Articles 37 and 38 of the Bulgarian and Romanian Accession Act). The former can be triggered if the Member State concerned causes a serious breach of the functioning of the internal market. The latter addresses serious shortcomings in the implementation of the AFSJ, such as the mutual recognition of judgments. In both cases, the Commission is empowered to adopt “appropriate measures”, which may even take the form of suspending the relevant rules of EU law in relation to the Member State concerned. This is a far-reaching power, expanding the Commission’s oversight well beyond Article 258 TFEU. The Cooperation and Verification Mechanism, adopted on this basis, illustrates the point.

Against this backdrop, three rather soft guidelines emerge from Article 49(2) TEU. In terms of scope, the notion of “adjustment” suggests minor changes that do not affect the overall structure of the Treaties (see Becker). This is supported by the change of wording at Maastricht, which shifted from “amendments”, “Änderungen”, or “révision” in Article 237(2) EEC to “adjustment”, “Anpassungen”, or “adaptations” in today’s Article 49(2) TEU (see Idriz, pp 175 ff). Today, “amendments” are reserved for Article 48 TEU. Formally, of course, any “adjustment” under Article 49(2) TEU is also an “amendment” of primary law. Still, the distinction between these two notions implies that the scope of the “change” that can be brought about under Article 49(2) TEU differs from that under Article 48 TEU. Admittedly, this line is elusive.

In terms of nexus, Article 49(2) TEU seems to require some link to accession. The provision refers to adjustments “which such admission entails”. Other language versions read even more narrowly: the adjustments must be “erforderlich” – necessary – or limited to those “que cette admission entraîne” – which result from the admission. In this spirit, the Court noted that “the integration of the new Member States … is the fundamental objective” of an accession act, which seeks to “facilitate their adjustment to the rules in force within the Community” (Commission v United Kingdom, 231/78, para 11). Whether this is the only permissible rationale, however, remains open. Ultimately, the nexus to accession can be determined by reference to admission requirements or political expediency. Member States enjoy broad political discretion here – there is no right to accede to the Union.

In terms of time, Article 49(2) TEU may require derogations to be of limited duration. In the aforementioned judgment, the Court stressed that the Act of Accession of Denmark, Ireland, and the UK “cannot be interpreted as having established for an indefinite period in favour of the new Member States a legal position different from that laid down by the Treaty for the original Member States … it would be contrary to the principle of the equality of the Member States … to accept that such inequalities could continue indefinitely” (ibid, para 17). Yet the Court interpreted only the act of accession, not its legal basis in Article 237(2) EEC, today’s Article 49(2) TEU. And indeed, the safeguard clauses in the Bulgarian and Romanian Accession Act were supposed to apply during the first three years after accession – “but they may be applicable beyond that date if the situation is still not remedied”. Accordingly, the Cooperation and Verification Mechanism remained in force from 2007 to 2023, more than five times as long as the original three-year window. Some derogations for acceding Member States are unlimited in time. Examples include Protocol No 6 limiting the acquisition of secondary residences in Malta by non-residents as an exception to the fundamental freedoms or the permanent safeguard clauses considered in the EU-Turkey negotiations.

The Commission’s proposal for an institutional safeguard clause seeks to stay within these guidelines by describing it as a “temporary, accession-specific mechanism for exceptional situations during the first years after accession.” Yet, the proposal clearly goes beyond the existing model in two respects. First, its scope is much broader, covering – beyond Article 2 TEU values – breaches of the principle of sincere cooperation. And second, the suggested measures are broader, including also the suspension of voting rights, which lie at the very heart of membership.

Limits: The Union’s Constitutional Framework

Early on, the existing safeguard clauses were criticised by Christophe Hillion as creating a “highly asymmetrical arrangement” that provokes a “double standard perception”.¹⁾ The proposed institutional safeguard clause takes these inequalities to another level. By allowing the suspension of voting rights in the Council under conditions different from those applying to “old” Member States, it would establish a form of second-class membership. Beyond the equality of Member States, this affects the equality of Union citizens. Under the Union’s structure of dual democratic legitimacy, Union citizens are represented directly in the European Parliament and indirectly through their Member State governments in the Council (Article 10(2) TEU). Different conditions for suspending Council voting rights therefore directly affect the equality of Union citizens in a central strand of democratic representation.

This raises the question of whether such an arrangement is constitutional. At first glance, one might object that Article 49(2) TEU provides a route to Treaty amendment. As acts of accession have the status of primary law, they may derogate from any Treaty provision. Several arguments, however, point in a different direction.

To start with, the Court has begun to distinguish a “constitutional framework” from the plethora of Treaty provisions, thus introducing a certain stratification of primary law. In this spirit, the Treaties can be read as expressing a structure of substantive hierarchy, in which some provisions carry particular interpretative weight. At the very core of this “constitutional framework” sits Article 2 TEU, but the equality of Member States in Article 4(2) TEU and the equality of Union citizens in Article 9 TEU could form part of it as well (see Klamert pp 86 ff). Along these lines, the Court has stressed that equality of Member States is intimately connected to solidarity, which forms part of the “fundamental basis of the EU legal order” (Commission v Hungary (Reclassification du cannabis), C-271/23, para 127).

Can such fundamental principles be altered by an “adjustment” of the Treaties under Article 49(2) TEU? Some scholars, such as Ulrich Becker or Christophe Hillion, have argued that accession treaties must leave the Union’s identity or core untouched. Interestingly, the Commission itself once took the view that derogations from primary law in accession treaties must be of limited duration, necessary for accession, and must not depart from the principles governing the Community (see its observations in Mattheus v Doego, 93/78, p 2208). The Court’s case law suggests limits too. In the aforementioned judgment on the UK’s accession act, the Court emphasised that its provisions “must be interpreted having regard to the foundations and the system of the Community, as established by the Treaty”, including the equality of the Member States (Commission v United Kingdom, 231/78, para 12). In its Opinion 1/91 on the EEA Agreement, the Court further held that association agreements concluded under what was then Article 238 EEC (today’s Article 217 TFEU) cannot set up a system of courts that conflicts with the “very foundations of the Community” (para 70). Importantly, “an amendment of Article 238 … could not cure the incompatibility with Community law” (para 71). Many – such as Carlo Curti Gialdino, Deirdre Curtin, or José Luis da Cruz Vilaça and Nuno Piçarra to name but a few – read this enigmatic passage as recognising a hard core of principles that cannot be amended.

One may construct or one may contest an unamendability doctrine in EU law. A narrower reading, however, understands this passage as establishing a transparency requirement: fundamental principles cannot be amended by the backdoor, through revision of other provisions, such as Article 238 EEC,²⁾ or – in our case – accession acts. If the Member States wish to revise the Union’s “very foundations” or its “constitutional framework”, they must do so openly and through the amendment procedure provided for in the Treaties – Article 48 TEU.

Alternatives: What, Then, Should the Union Do?

If the problem is the ability of a single Member State to block the Union, the answer must be a general one: unanimity should give way to qualified majority voting – the passerelle clauses already provide for that option. The remaining “choke points”, especially in the CFSP, can be addressed through interpretation, drawing on the adaptations advanced by the Commission (pp 4 f) or on scholarly proposals, some more, some less controversial. If the problem is the protection of the Union’s values, the answer must likewise be a general one: the existing toolbox should be used and, where necessary, strengthened – this is possible under the current Treaty framework as well. Whether an accession treaty could introduce institutional safeguard clauses applying to all Member States – as an “adjustment” under Article 49(2) TEU – remains an open question. The equality of Member States and of Union citizens, at least, would not stand in the way. And where the Member States want more, there is Article 48 TEU.

The temptation to escape the dilemma of enlargement without reform by resorting to accession treaties is understandable. Rules written for others are always easier to agree on than rules written for ourselves. But enlargement has always been a mirror. The question it puts to us is not only whether Ukraine, Albania, Montenegro, or Moldova are ready for the Union but also whether the Union is ready to receive them as equals. Institutional safeguard clauses that create deep inequalities among Member States and Union citizens are not the way.

The author wishes to thank Christophe Hillion for inspiring discussions on the topic.