AN CHÚIRT UACHTARACH
THE SUPREME COURT
S:AP:IE:2025:000148
[2026] IESC 27
O’Donnell C.J.
O’Malley J.
Hogan J.
Murray J.
Collins J.
IN THE MATTER OF SECTIONS 142 AND 150 OF THE
DATA PROTECTION ACT 2018
BETWEEN/
TIKTOK TECHNOLOGY LIMITED AND TIKTOK INFORMATION
TECHNOLOGIES UK LIMITED
APPLICANTS/
RESPONDENTS
AND
DATA PROTECTION COMMISSION
RESPONDENT/
APPELLANT
JUDGMENT of Mr. Justice Gerard Hogan delivered on the 30th day of April 2026
1
Background
1. The concept of national procedural autonomy is a fundamental principle of European
Union law. It reflects the very practical reality that not only is the Union composed of
different types of legal families, but also that each of the Member States have their own
legal systems and their own very different procedures. This national procedural
autonomy is, of course, tempered by the concepts of equivalence and effectiveness
which the Court of Justice has developed over the years: see, e.g., Lenaerts, Gutman,
Nowak eds., EU Procedural Law (Oxford, 2023) at pp. 122-124. There may, of course,
be special cases where, exceptionally, Union law reaches into national procedural law
and specifies that a particular remedy must be made available. These cases are,
however, exceptional and, reflecting this fact, where the Union legislature so acts, it
does so in express terms.
2. Over and above all of this, the concept reflects an even more profound principle,
namely, the division of juridical sovereignty as between the Union and its Member
States. Just as the Member States must uphold the supremacy of Union law within its
proper sphere of application by virtue of their duty of loyal co-operation under Article
4(3) TEU, so too must the Union respect the autonomy of the Member States. After all,
as the German Constitutional Court observed in its landmark decision of 12th October
1993 in the Maastricht Treaty case (89 Bverf GE 150 at 190), it is the Member States
who are the masters of the Treaties (“Herren der Verträge”) and not the other way
around. The Union does not, therefore, have its own autochthonous sovereignty. It
instead rather enjoys only that with which it has been conferred by the Member States.
Subject to the exceptions I have mentioned, national procedural autonomy remains the
preserve of the Member States. It has never been transferred to the Union.
2
3. All of this means that the starting point for this appeal is that the rules and practice in
relation to the grant of a stay (in effect, a form of injunction) are governed by Irish law,
subject only to the principles of equivalence and effectiveness. The Data Protection
Commission (“DPC”) argues, however, that the power to grant a stay should be
governed by EU law, specifically by reference to the judgment of the Court of Justice
in Joined Cases C-143/88 and C-92/89 Zuckerfabrik Suderdithmarschen AG
EU:C:1991:65. I shall return presently to consider this judgment and the subsequent
application of these principles in the recent case-law.
4. For the moment, however, it suffices to record that the present application for a stay is
in form an application to restrain the operation of an administrative decision of the DPC.
But this is an administrative decision which, as Murray J. will explain in the judgment
he is about to deliver, will have far-reaching consequences for the operation of the
applicant’s business model. TikTok have claimed that its financial losses will exceed
€1bn, if no stay is granted. In addition, the impugned decision might well require
thousands of TikTok employees to re-locate from their present place of employment in
the People’s Republic of China to a variety of other jurisdictions in Southeast Asia, the
United States and elsewhere. For its part the DPC counters this argument by saying that
the privacy rights of millions of TikTok consumers would be potentially compromised
if a stay on its decision is granted.
5. Against that background, I propose first to examine the jurisdiction under our national
procedural law to grant a stay in respect of the operation of administrative decisions. I
will then consider whether a stay ought to have been granted in the present case. Finally,
I will address the question of whether in this instance national procedural law has been
ousted by EU law, specifically the Zuckerfabrik test.
3
The jurisdiction to grant a stay in respect of the operation of an administrative
decision
6. Turning first to the jurisdiction to grant a stay under our national procedural law, one
may observe that our law has long provided for the power to grant a stay or an injunctive
type remedy of an interim or interlocutory character. Many applications for a stay of
this kind arise in the context of judicial review applications where this procedure is
nowadays regulated by the provisions of Ord. 84, r. 20(8) of the Rules of the Superior
Courts. Yet over and above the provisions of Ord. 84, our national procedural law has
also provided for the power to grant a stay or an injunctive type remedy of an interim
or interlocutory character. This was the position at common law in 1937 at the time of
the adoption of the Constitution and these procedural rules and practices were carried
over by Article 50 of the Constitution. These rules were in turn augmented by the
provisions of Article 40.3.2⁰ of the Constitution which commit the State to ensuring
that an effective remedy will be provided to litigants: see, e.g., Efe v. Minister for
Justice [2011] IEHC 214, [2011] 2 IR 798; NM (DRC) v. Minister for Justice [2016]
IECA 217, [2018] 2 IR 591.
7. In saying, therefore, that the Constitution ensures that, subject to considerations of
practicability, a litigant will be granted an effective remedy, one is doing no more than
giving effect to the express language of Article 40.3.2⁰ (“…protect as best it may from
unjust attack and, in the case of injustice done, vindicate the life, person, good name,
and property rights of every citizen”.) There is no need to characterise this as some
form of “derivative” constitutional right in the sense understood by our decision in
Friends of the Irish Environment v. Government of Ireland [2020] IESC 49, [2021] 3
IR 1: it is rather instead a case of giving effect to what the actual text of the Constitution
simply ordains, even if Article 40.3 does not in terms use the phrase “effective remedy”.
4
8. Irrespective of how the present application for a stay is described or formulated, it is
clear the present proceedings involve a challenge to an administrative decision with far-
reaching implications for a large number of third parties, including thousands of TikTok
employees and millions of consumers of this social media service. Perhaps just as
importantly, as Mulcahy J. acknowledged (at para. 148 of his judgment) in the High
Court, as the DPC’s decision “will have binding effect on most, if not all, [other EEA]
supervis[ory] authorities”, a stay on that decision will have considerable practical
implications across the EU/EEA.
9. To that extent, the present application for a stay has implications which go beyond the
standard inter partes application for an interlocutory injunction. This is, of course, true
of many applications for stays of regulatory decisions of this kind. Where (as here) the
regulatory decision in question has implications for other supervisory authorities, the
wider industry and the public at large, then such an application for a stay may be
regarded as closer to those cases which involve an application to stay the operation of
a legislative instrument, i.e., classically an application for an injunction to restrain the
operation of a statute pending the outcome of a constitutional challenge.
10. This Court has recognised as far back as 1960 the existence of a jurisdiction to grant
relief pending the outcome of a constitutional challenge: see The State (Attorney
General) v. Mangan (1961) Irish Jurist Reports 17. In that case this Court held that the
District Court was entitled to adjourn a criminal prosecution pending the outcome of a
challenge to the constitutionality of the legislative provision providing for that offence.
While the judgment of Maguire C.J. admittedly did not address in terms the question
of granting interim relief, it nonetheless certainly hints in that direction. Maguire C.J.
thus stated (at 33):
5
“It has to be remembered that the enactment of a written Constitution created
an entirely new situation…The legislature is no longer supreme. Its laws to be
valid also must be in accordance with the provisions of the Constitution. This
new position must be recognised and accepted by everybody and is the
particular concern of those who are called upon to exercise judicial functions.”
11. The first formal recognition by this Court of a jurisdiction to grant an injunction to
restrain the operation of an allegedly unconstitutional statute came, however, with our
decision in Pesca Valentia Ltd. v. Minister for Fisheries [1985] IR 193. As Finlay C.J.
observed ([1985] IR 193 at 201):
“It is, as has been so frequently stated, the duty of the courts to protect persons
against the invasion of their constitutional rights or against unconstitutional
action. It would seem wholly inconsistent with that duty if the Court were to be
without power in an appropriate case to restrain by injunction an action against
a person which found its authority in a statutory provision which might
eventually be held to be invalid having regard to the Constitution.”
12. One cannot but agree with this statement of principle. If it were otherwise, it would
mean, for example, that the courts would have to stand by, helpless to assist the plight
of citizens whose rights might be affected by even an egregiously unconstitutional law
pending a formal determination as to its constitutional validity. Such a state of affairs
would be totally inconsistent with this Court’s general constitutional duty under Article
40.3.2⁰ to “protect as best it may from unjust attack and, in the case of injustice done,
vindicate the life, person, good name, and property rights of every citizen”.
13. While the existence of this general jurisdiction is not in doubt, the difficulty which
arises concerns the application of these principles. While Finlay C.J. said in Pesca
Valentia ([1985] IR 193 at 201) that there was no “special principle applicable to an
6
application for an interlocutory injunction of this kind”, this statement must, I think, be
viewed with some reserve and as one which in fact has in practice been overtaken by
the subsequent case-law. Where a plaintiff seeks an interlocutory injunction in aid of a
constitutional challenge (or, for that matter, where - as here – there is a challenge to the
validity of a regulatory decision with far-reaching implications for the industry and
consumers alike) different factors generally come into play as compared with, for
example, the standard interlocutory injunction application in a commercial dispute. In
the latter case the court is required only to balance the rights of the parties inter se,
whereas in a constitutional challenge the court also has to have regard to the wider
public interest, including the interest in upholding the effective operation of the law.
14. This was the very point which Clarke J. made in the High Court in MD (an infant) v.
Ireland [2009] IEHC 206, [2009] 3 IR 690. That was a case where an underage male
was charged with having sexual intercourse with an underage female. Section 3(1) of
the Criminal Law (Sexual Offences) Act 2006 had provided that it was an offence for
an underage male to engage in such conduct with an underage female, but by virtue of
the provisions of s. 5, the converse was not true. The plaintiff sought a stay on the
prosecution pending the outcome of the constitutional challenge.
15. Clarke J. first referred to Pesca Valentia and then continued ([2009] 3 IR 690 at 694-
695):
“. … it was argued, correctly in my view, that the relevant jurisdiction is one
which must be most sparingly exercised. The reasons for this are obvious.
Legislation which has been passed into law by the Oireachtas enjoys a
presumption of constitutionality. If it were to be the case that persons who were
able to establish a fair case to be tried concerning the validity of the relevant
legislation having regard to the provisions of the Constitution (which is not a
7
particularly high threshold) were able to obtain an injunction preventing, in
practice, the application of the legislation to them until the proceedings had been
determined, then it would follow that legislation could, in practice, be sterilised
pending a final determination of the constitutional issues raised. Those
considerations apply with equal force where the statute concerned is one which
creates a criminal offence.
While, in general terms, the principles applicable to the grant or refusal of an
interlocutory injunction in a case such as this are no different from those which
apply in the case of any other interlocutory injunction, it has to be emphasised
that a very significant weight indeed needs to be attached, in considering the
balance of convenience, to the desirability that legislation once coming into
force should be applied unless and until such legislation is found to be invalid
having regard to the Constitution. It should only be where significant
countervailing factors can be identified or where it is possible to put in place
measures which would minimise the extent to which there would be any
interference with the proper and orderly implementation of the legislation
concerned, that a court should be prepared to grant an injunction which would
have the effect of preventing legislation which is prima facie valid from being
enforced in the ordinary way.” (emphasis supplied)
16. While Clarke J. did grant a stay in respect of the criminal prosecution in that case, this
was limited in terms and subject to strict conditions. The key point is that there was a
complete judicial awareness of the potential implications of the grant of a stay and how
the effectiveness of the legislation was a factor to which the court must have regard.
17. Clarke J. made the same point for this Court in his judgment in Okunade v. Minister
for Justice [2012] IESC 49, [2012] 3 IR 152. Here the applicants applied for an
8
interlocutory injunction restraining the operation of a deportation order pending the
outcome of a challenge to its validity. While this was different from both Pesca Valentia
and MD in that it was not, as such, a constitutional challenge, Clarke J. nonetheless
stressed ([2012] 3 IR 152 at 188-189) that there was “a further feature of judicial review
proceedings which is rarely present in ordinary injunctive proceedings.” He continued
(at 188):
“The entitlement of those who are given statutory or other power and authority
so as to conduct specified types of legally binding decision making or action
taking is an important part of the structure of a legal order based on the rule of
law. Recognising the entitlement of such persons or bodies to carry out their
remit without undue interference is an important feature of any balancing
exercise”.
18. If, for example, an interlocutory injunction is granted restraining the operation of a
statute it may have the practical effect in some instances of placing that statute into a
form of suspense, perhaps for months or even years. As Clarke J. observed in both MD
(in the High Court) and Okunade (in this Court), this would have serious implications
for the rule of law. This was the very point which had previously been made by Henchy
J. in The State (Llewellyn) v. Ua Donnchadha [1973] IR 151 at 158, where he expressed
the concern that if such relief could be routinely granted “all that would be necessary
to bring every prosecution in the District Court under a particular statutory provision to
a stop would be to get a nominal plaintiff to institute proceedings in the High Court to
have the statutory provision declared unconstitutional.” Similar concerns were
expressed by Murphy J. in Nova Media Services Ltd. v. Minister for Posts & Telegraphs
[1984] ILRM 161 at 168, when he spoke of the “damage which might be caused to the
common good by the interference with statutory controls and regulations [which might]
9
be very considerable indeed and certainly…could never be assessed in monetary
terms.”
19. All of this means that the exercise of the Pesca Valentia jurisdiction is generally
confined to special or unusual cases, such as where the underlying constitutional
challenge can be regarded as powerful and strong. While the jurisdiction is – to adopt
the language of Clarke J. in MD – “sparingly exercised”, I consider, however, that the
dictum of Murphy J. in Nova Media Services to the effect that such relief would only
be granted in “the most extraordinary circumstances”, places the test at too elevated a
level. This statement has, in any event, been overtaken by what was subsequently said
by this Court in Pesca Valentia.
20. To that extent Irish procedural law conforms fully with the general principle of
Zuckerfabrik in that this Court clearly enjoys a jurisdiction in an appropriate case to
grant an interlocutory order restraining the operation of a statute which was said to be
unconstitutional or, for that matter, contrary to EU law. That, indeed, was the
background to Pesca Valentia, where the plaintiff shipowners claimed that the
nationality criterion in respect of crew members which had been imposed by the
Fisheries (Amendment) Act 1983 was unconstitutional and contrary to EU law.
Whether a stay ought to be granted by reference to national procedural law
21. As this Court has made clear in a range of decisions – from Pine Valley Developments
Ltd. v. Minister for the Environment [1987] IR 23 through to Barlow v. Minister for
Communications [2025] IESC 14 – the law recognises that in the modern administrative
state there will inevitably be categories of unconstitutional and unlawful administrative
actions causing loss and damages for which there is either only limited or, as often as
not, no compensation or redress. There are, to that extent, categories of damnum sine
injuria. But while the law recognises the existence of damnum sine injuria, it takes, so
10
to speak, no pleasure from it and it seeks to avoid this where possible. This is reflected
in the manner in which the balance of convenience is assessed, since the courts will
have regard to this factor and the practical availability of an action for damages in
considering whether damages are an adequate remedy: see, e.g., Dowling v. Minister
for Finance [2013] IESC 37, [2013] 4 IR 576 and Word Perfect Translation Services
Ltd. v. Minister for Public Expenditure and Reform [2018] IECA 35, [2019] 2 IR 503.
22. It is true that in MD (a minor) v. Board of Management of a Secondary School [2024]
IESC 11 we reserved the question of whether the High Court enjoyed a jurisdiction to
grant an injunction in aid of a challenge to the validity of an administrative decision.
That, however, was in a context where the only appeal lay from a decision of a school
board to an administrative panel established by the Education Act 1998. (This is a
different MD case to the one in which Clarke J. gave judgment in the High Court.)
23. In the present case, however, it is unnecessary to pronounce on that wider question.
The traditional view was that the remedy followed the right, so that no injunction could
be granted in those cases where the courts had no underlying jurisdiction in respect of
the dispute. It may well be, however, that the overriding obligation imposed on the
courts by Article 40.3 to provide for an effective remedy in respect of challenges to
administrative decisions will to some degree oblige the courts to depart from the
traditional practice whereby an injunction could only be granted in aid of a substantive
right.
24. In any event, it is not necessary to decide this question, since by contrast with the
background of MD (where the appeal against the suspension of a school student lay to
an administrative body), the Oireachtas has provided for a statutory appeal to the High
Court under s. 150 of the Data Protection Act 2018 (“the 2018 Act”) against a decision
of the DPC. While it is also the case that the 2018 Act does not in terms provide for the
11
grant of interim relief, this omission is not material. Given that the High Court enjoys
a full original jurisdiction under Article 34.3.1⁰ of the Constitution with “power to
determine all matters and questions whether of law or fact”, it follows that it must also
enjoy a jurisdiction to grant interim relief in respect of the administrative decision under
challenge. As I have already noted, the plain language of Article 40.3.2⁰ demands no
less.
25. For good measure one might in any event note that s. 150 of the 2018 Act has the
heading of the words “Right to effective judicial remedy”. This reflects the language of
Article 78(1) of the General Data Protection Regulation (EU) 2016/679 of 27th April
2016 of the European Parliament and Council (“the GDPR”) itself. (The Court is, I
think, entitled to have regard to the language of this statutory heading in the light of the
provisions of s. 7 and s. 18(g)(ii) of the Interpretation Act 2005). It suffices to say here
that if the High Court could not grant interim relief in aid of the statutory appeal, TikTok
would be denied a key aspect of an effective judicial remedy in the manner guaranteed
by the GDPR.
26. To that extent, therefore, so far as the present appeal is concerned, TikTok is in principle
entitled to apply for a stay so as to ensure that its constitutional rights (specifically, its
property rights in its assets) are protected pending the outcome of its statutory appeal
to the High Court. For all the reasons which Murray J. will elaborate in his judgment, I
consider that TikTok should be granted interim relief in aid of its statutory appeal. It is
accepted that TikTok has raised weighty matters in its statutory appeal and one may
note that, as of the date of preparing this judgment, this appeal has already been heard
by the High Court and judgment is currently awaited. In these circumstances it suffices
for me to point to factors such as the relatively short duration of the stay and the fact
that TikTok are likely to incur huge costs if no stay is granted. In weighing these factors,
12
one cannot overlook the fact that these costs, once incurred, are likely to be
irrecoverable, even if TikTok were to win that appeal.
Has national procedural law been supplanted by EU law so far as the jurisdiction
to grant a stay is concerned?
27. Turning to the question of whether national procedural law has, in this respect, been
supplanted by EU law, it is important to note that the DPC accepts that the Okunade
test does not, as such, violate the principle of effectiveness. The DPC also accepts that
the statutory appeal will itself be determined by reference to national law. This is, after
all, what is contemplated by Article 78(3) of the GDPR which provides:
“Proceedings against a supervisory authority shall be brought before the courts
of the Member State where the supervisory authority is established”.
28. As Mulcahy J. observed in his judgment (at paragraph 155), Recital 143 of the GDPR
further provides that such proceedings “should be conducted in accordance with that
Member State’s procedural law. Those courts should exercise full jurisdiction, which
should include jurisdiction to examine all questions of fact and law relevant to the
dispute before them”. (Emphasis supplied).
29. One might have thought that this would be sufficient to dispose of the DPC’s arguments
in this regard. After all, this Court confirmed in Dowling that the Okunade test satisfied
the requirements of equivalence and effectiveness and that the Zuckerfabrik test applied
only where there is a challenge to the legality of an EU legislative act.
30. The DPC ripostes, however, by advancing two distinct arguments. First, it says that the
decision under challenge was made jointly with other supervisory authorities concerned
(“CSAs”) in this matter. Second, it says that a national procedural law test would
undermine the full effectiveness of EU law («pleine efficacité»). We can now examine
these arguments in turn.
13
Was the decision of the DPC a “joint decision” made with other supervisory
authorities?
31. It is true that the GDPR provides for extensive consultation between all the supervisory
authorities of the Member States. But as Mulcahy J. correctly noted in his judgment,
the decision of the lead supervisory authority (“LSA”)(i.e., the DPC) is a decision of
that body. I think that counsel for TikTok was correct to emphasise that any other
conclusion would be inconsistent with Article 78 of the GDPR. Article 78(1) clearly
provides for a judicial remedy in respect of “a legally binding decision of a supervisory
authority.”
32. It is true that, as counsel for the DPC observed, recital 126 provides that the decision
should be agreed jointly by the LSA and the CSAs. Yet it is clear from Article 60(7)
that it is the LSA which makes the final decision, a conclusion which I think is at least
necessarily implicit in the judgment of Anthony Collins J. in Meta Platforms Ireland
Ltd. v. Data Protection Commission [2025] IECA 60. This, in any event, was confirmed
by the Court of Justice in Facebook Ireland Ltd. (Case C-645/19, EU:C: 2021:483),
subject to two limited exceptions not material to the present case. As that Court noted
(at paras. 51-56), while the LSA cannot ignore the views of the other CSAs and while
the various reasoned objections of the CSAs may have the effect of blocking the
ultimate decision pending its resolution, the decision which is adopted remains that of
the LSA.
33. A variant of this argument relates to the role of the European Data Protection Board
(“EDPB”) in the dispute resolution process. Article 65(1)(a) of the GDPR provides that
in such circumstances the EDPB decision on the matters in dispute will be binding. Yet
the final decision remains in law that of the LSA, even if that decision must also refer
to the EDPB decision and must follow it: see C-97/23P WhatsApp Ireland Ltd.
14
EU:C:2026:81, at para. 71. It follows, therefore, that the “joint decision” argument
cannot be accepted.
Whether the national law test in relation to stays would violate the general
effectiveness of EU law
34. The DPC contends, however, that the test posited by our national law in respect of stays
on regulatory decisions of this kind would undermine the general effectiveness of EU
law and that it should be supplanted by the Zuckerfabrik test. It draws attention to the
fact that stays granted by national courts can frustrate or even impede regulatory
objectives and measures. To the end, counsel for the DPC placed emphasis on the
decision of the Vice-President of the Court of Justice in Case C-639/23P(R) Amazon
Services Europe Sárl EU:C:2024:277 where he stated (at para. 157) that “…not
applying certain obligations laid down by [the Digital Services Act, Regulation (EU)
2022/2065] will lead to a delay, potentially for several years, in the full achievement of
those objectives. Not applying those obligations will therefore give rise to a risk of
potentially allowing an online environment which threatens the fundamental rights
provided for in the Charter to persist and develop”.
35. These points are all well made by counsel for the DPC, but they must be placed in
context. First, it is true that as the Vice-President observed in Amazon Services, if an
important regulatory decision was stayed for several years, this could compromise the
full objective of these regulatory decisions. In the present case, however, there is no
risk of this, since the duration of the stay is to be measured in weeks and at most a few
months. Second, those comments of the Vice-President were made in the context of a
stay which would effectively have disapplied the entirety of the Digital Services Act in
the case of a major on-line platform such as Amazon. While I do not seek to minimise
the potential impact on the privacy of consumers by reason of TikTok’s transfer of EEA
15
user data to the People’s Republic of China, this has to be weighed against the
potentially enormous costs which TikTok will very probably suffer if no stay is granted.
This is, in essence, why I would favour the grant of a stay for the limited time duration,
i.e., pending the outcome of the statutory appeal to the High Court. (At the time of
writing the High Court had already reserved judgment in respect of that statutory
appeal).
36. Third, in any event, the Pesca Valentia test must now be read in the light of a series of
subsequent decisions, not least by reference to cases such as MD and Okunade. There
is a complete judicial awareness of the potential implications of the grant of a stay in
respect of regulatory decisions. To that extent, there is really very little difference in
principle between the Zuckerfabrik and Okunade tests.
37. The other argument is that the full effectiveness of EU law requires that national
procedural law be supplanted by a version of the Zuckerfabrik test. The straightforward
response to this argument is that there is nothing in the GDPR which provides for this.
Quite the contrary: the Union legislature plainly intended that national procedural law
would apply. If, for example, the final decision must be taken by the LSA, it follows
that the validity of that decision is governed by national law. It would be strange – and
internally inconsistent with the structure of the GDPR – if one particular feature of
national procedural law alone (namely, the power to stay a decision of the LSA) was to
be supplanted by EU law, while leaving the result of national law intact and unaffected.
38. The furthest that this argument was taken was the suggestion - and it really was no more
than that – that the Okunade test was inadequate to safeguard the full effectiveness of
EU law. That might possibly be the case if the test meant that the courts granted
something in the nature of an automatic stay such that the full effect of EU law was
undermined or that a regulatory regime was de facto placed into abeyance for several
16
years. This is, however, not what Okunade says and under that test the national courts
will (and must) take full cognisance of these wider public interest considerations,
including of course the obligation to apply EU law. There is, in any event, no Court of
Justice decision which goes further than requiring that national procedural law must
conform to the principles of equivalence and effectiveness.
39. It follows, therefore, that for those reasons I would reject the DPC’s arguments on this
front.
Whether this Court should make an Article 267(3) TFEU reference
40. This Court is, of course, a court of last resort for the purposes of Article 267(3) TFEU.
Accordingly, in view of the decision of the Court of Justice in Consorzio Italian
Management (Case C-561/19, EU:C:2021:799) and the comments of that Court (at
para. 51 of the judgment) regarding the nature of the obligation to refer which is
imposed on courts of last resort, it is appropriate to record why I do not think it would
have been necessary to make a reference of any question of the interpretation of EU
law to that Court.
41. While it is true that the present application concerns an interlocutory matter, I think that
counsel for the DPC was correct when she submitted that the obligation to refer under
Article 267(3) TFEU still applied where the interlocutory issue would not itself be
revisited during the course of the main hearing. The issue of EU law which potentially
arises on this appeal relates intrinsically to the interlocutory issue only, i.e., the test for
a stay. As this issue will not be re-opened at a later stage of the proceedings, I agree
that Article 267(3) TFEU is in principle applicable.
42. The relevant test here was set out in Hoffmann-La Roche (Case C-107/76; EU:C:1977:
89) where the Court of Justice stated that Article 267(3):
17
“… must be interpreted as meaning that a national court or tribunal is not
required to refer to the court a question of interpretation or of validity mentioned
in that article when the question is raised in interlocutory proceedings for an
interim order ….even where no judicial remedy is available against the decision
to be taken in the context of those proceedings provided that each of the parties
is entitled to institute proceedings or to require proceedings to be instituted on
the substance of the case and that during such proceedings the question
provisionally decided in the summary proceedings may be re-examined and
may be the subject of a reference to the court under [Article 267].”
43. While this is so, the Court of Justice has also confirmed that it has no jurisdiction to
entertain the reference where the interlocutory orders in question have already expired
or have been overtaken by events. Thus, in Case C-338/85 Fratelli Pardini SpA
EU:C:1988:194 the Court of Justice held (at para. 11) that a national court was:
“not empowered to bring a matter before the Court by way of a reference for a
preliminary ruling unless a dispute is pending before it in the context of which
it is called upon to give a decision capable of taking into account the preliminary
ruling. Conversely, the Court of Justice has no jurisdiction to hear a reference
for a preliminary ruling when at the time it is made the procedure before the
Court making it has already been terminated”. (emphasis supplied)
44. This was the approach taken by this Court in Dowling where we decided not to make a
preliminary reference because, as Clarke J. explained, “the questions are raised in
interlocutory proceedings and would be overtaken by events in any case”. I think that
the present case is essentially similar. The original stay was designed to last pending
the full hearing of the statutory appeal. But that hearing has already been heard before
18
the High Court and judgment is currently awaited. Any decision of that Court regarding
some future stay will be governed by new and different considerations once that hearing
is completed.
Conclusions
45. For the reasons stated, I accordingly join with the judgment of Murray J. I agree that
the stay should be continued for the relatively short period involved pending the
outcome of the High Court hearing. It is not disputed that TikTok can point to the
existence of substantial grounds in support of their statutory appeal. It is equally clear
that TikTok are likely to suffer substantial losses which would, in practice, be
effectively irrecoverable if no stay were granted. Given, moreover, that the stay will
last for a relatively short period pending the delivery of the judgment of the High Court,
all of this points in favour of the grant of a stay.
46. Given that the initial stay will have been overtaken by events, there is in the wake of
the Court of Justice’s judgment in Fratelli Pardini no obligation to refer any question
to the Court of Justice under Article 267(3) in relation to the operation of the stay. In
any event, I see no essential difference between Zuckerfabrik and Okunade so far as the
relevant tests for the grant of a stay on the operation of an administrative decision is
concerned.
47. I would accordingly dismiss the appeal.
19