Following a catastrophic administrative error placing the lives of thousands of Afghans at risk, the UK government obtained a super-injunction, enabling it to attempt to clear up the mess it had created behind a veil of impenetrable secrecy. The courts’ willingness to keep the super-injunction in place for nearly two years raises some important constitutional questions.
The background to the super-injunction granted to prevent both the discloure of a highly sensitive dataset containing the details of tens of thousands of Afghans who had assisted the UK government (and, it now transpires, some UK intelligence and security personnel too) and the disclosure of the injunction’s very existence is by now well known. It is therefore sufficient for present purposes to rely on the succinct statement of the facts provided by Chamberlain J in paragraph 6 of the first of the four judgments that he issued in this case:
It is not necessary to set out here the circumstances in which the compromise occurred, save that it appears to have involved an error by an individual who works for the UK Government. The dataset included the identities of many thousands of individuals who applied to come to the UK under the Afghanistan Relocations and Assistance Policy (“ARAP”), generally because of their past work for, or connection to, the UK. If these data were to fall into the hands of the Taliban, who now control the whole of Afghanistan, the individuals in question and their families would be at risk of being killed or subject to serious harm. The MOD [Ministry of Defence] fears that, if the existence of the data breach becomes known to the Taliban, they will take steps to acquire the dataset and may be able to do so.
A super-injunction contra mundum (that is, against the whole world) was granted by Robin Knowles J on 1 September 2023, even though the government had requested only an injunction (to prevent disclosure of the dataset) as distinct from a super-injunction (which would prevent disclosure of the existence of the injunction as well). In his first judgment, Chamberlain J explained the extraordinary implications of such injunctions — ‘the court is using its coercive powers to curtail the right of would-be publishers to freedom of expression and the correlative right of the public to receive information — and doing so in circumstances which are entirely insulated from public scrutiny’ — and the corresponding need for super-injunctions to be ‘justified by the most compelling evidence’ and issued ‘only in truly exceptional circumstances’.
That Robin Knowles J and subsequently Chamberlain J (and, as we will see, the Court of Appeal) were confronted with a difficult problem is beyond doubt. On the one hand, it was argued that the lives of thousands of Afghans would be placed at risk if the existence and content of the dataset became known to the Taliban. On the other hand, the implications of such a super-injunction — not least in terms of freedom of expression, the capacity of the media to perform its essential role and the ability of Parliament to scrutinise government decision-making and policy — would be hard to overstate.
In his second judgment, Chamberlain J agreed in February 2024 to the continuation of the super-injunction, but held in his third judgment in May 2024 that the injunction should be discharged. That judgment was overturned by the Court of Appeal, the injunction remaining in force until July 2025, when Chamberlain J, in his fourth judgment, discharged it, nearly two years after it had been granted.
Scrutiny of the government’s justification
From a constitutional perspective, two aspects of this series of cases strike me as particularly interesting and important. The first concerns the approach taken by the courts to scrutinising the justification offered by the government in support of the granting and continuation of the injunction. Bearing in mind that the issues at stake ultimately turned upon assessments of what was likely to happen if the super-injunction was not in place, thereby risking the Taliban’s becoming aware of the existence and potentially the content of the dataset, Chamberlain J explained in paragraph 35 of his first judgment the overarching principles to be applied by the court:
(a) The predictive judgment of the executive about the extent of any risk is entitled to respect.
(b) However, it is for the court, not the executive, to decide where the balance of public and private interests falls.
(c) It is not enough for the Government to show that the disclosure of information will give rise to a risk of life … [T]he Court also must also consider whether the grant of relief might endanger life. If so, it is for the Court to consider and balance these consequences.
(d) The court’s decision on any interim order must take into account both this balance of private interests and the public interest in open justice and in the transparency of public decision-making.
He went on, at paragraph 36ff, to identify important reasons why ‘these latter public interests’ — ie the interests in open justice and transparency — ‘have a special significance and weight’ in the context of the present case. He observed that super-injunctions are inevitably ‘interferences with freedom of expression which take place under the radar’, with the public remaining wholly ignorant of them, and particular concerns arising when the applicant is the Government, given that the ‘grant of a super-injunction to the Government is likely to give rise to understandable suspicion that the court’s processes are being used for the purposes of censorship. This is corrosive of the public’s trust in Government.’ Chamberlain J also pointed out that a great many decisions — about whom to admit to the UK and the extent of efforts to ameliorate the potential impact of the data breach — would need to be taken. While the media and Parliament would normally ensure that the Government was ‘subject to the ordinary mechanisms of accountability which operate in a democracy’, ‘[t]he grant of a super-injunction has the effect of completely shutting down these mechanisms of accountability, at least while the injunction is in force.’ It followed that the maintenance of the super-injunction would cause such decisions to be taken in a ‘scrutiny vacuum’, which compromised the axioms that ‘decisions subject to public and Parliamentary scrutiny are not only more legitimate, but are also likely to be better than ones taken in secret’.
These are important points, well made. And while the context is different — this was not, after all, a judicial review case in which the lawfulness of a government decision was being impugned — the approach commended by Chamberlain J finds plenty of support in contemporary case law recognising that courts do not need to, indeed should not, adopt an unduly deferential approach when it comes to evaluating justifications offered by the Government in circumstances in which important rights or constitutional values are in play. As explained above, Chamberlain J was prepared to accept the Government’s justification and to continue the injunction in his first and second judgments, but his position had shifted by the time of his third judgment in May 2024. He explained at paragraph 18 of his third judgment that:
At each of the hearings before me, the application for relief has been advanced on the basis that extra time was required to undertake further work to mitigate the effect of the data breach; and the expectation has been that the justification for the injunction would likely fall away in a matter of weeks, or at most a few months, because the Taliban were likely to find out about the data breach and then gain access to the dataset. This eventuality was termed a “break glass” event. The MOD assessment that the Taliban have not learned about the existence of the data breach, and have not obtained access to the dataset, has remained unchanged throughout. I shall return to this point later. The important point for present purposes, however, is that, on the MOD’s current case, there has not been a “break glass” event for some 8 months, and it must follow that the injunction may well be required for a very substantial further period.
He said (at paragraphs 27–28) that ‘the nettle must therefore now be grasped’ because ‘the point has now been reached where a decision to continue the injunction would logically entail continuing it for a very considerable further period, unless and until evidence of a “break glass” event emerged’. His decision to discharge the injunction was informed by his assessment that by May 2024, there was a ‘significant possibility’ that the Taliban knew of the dataset’s existence, in which case both the several thousand Afghans slated for relocation to the UK and the tens of thousands of individuals mentioned in the dataset who were not set to be moved might be placed at increased risk by the continuation of the super-injunction, given that it would prevent them from being aware of the risk they faced and taking steps to mitigate it. Moreover, the non-relocation cohort would be denied the possibility of seeking to apply pressure to the UK authorities to include them in the relocation plans. Alongside such considerations, Chamberlain J placed emphasis (at paragraph 47) on the importance of public and political scrutiny:
The size and import of the decisions now being taken, both in moral and in financial terms, bring this point into sharp relief. It is fundamentally objectionable for decisions that affect the lives and safety of thousands of human beings, and involve the commitment of billions of pounds of public money, to be taken in circumstances where they are completely insulated from public debate.
Against that background, he concluded that the injunction should be discharged. The Court of Appeal, however, disagreed, for reasons that Chamberlain J himself helpfully summarised in his fourth judgment:
The MOD appealed to the Court of Appeal. That court decided … that I had made two main errors. First, I had failed to compare what would happen if the injunction were discharged with what would happen if it were continued. “On the undisputed evidence,” they concluded, “the effect of discharging the injunction would be to turn what was a real possibility into a virtual certainty”: [61]. This was a reference to the Taliban acquiring the dataset. Secondly, there was an error in relation to the finding of an adverse effect on the non-relocation cohort. It was wrong to consider that public pressure might assist that cohort, because there would not be time for that pressure to bear fruit in the form of a change in policy: see CA Judgment, [59]-[65]. The Court of Appeal went on to conclude that there was no material change from the position in my Judgment nos 1 and 2 and that the super-injunction should be continued, subject to periodic review by me at least every three months, noting that “the usefulness of the dataset to hostile actors is likely to reduce over time”: see CA Judgment, [80]-[81].
With respect to the Court of Appeal, it seems to me that its judgment falls into precisely the type of error that Chamberlain J cautioned against in his first judgment. While the Government’s predictive judgment about the extent of relevant risks might be due an appropriate degree of ‘respect’, it is for the court itself to determine the relative weight to be accorded to the private and public interests in play. The Court of Appeal took Chamberlain J to have concluded that the super-injunction should be discharged simply because it had been in place for as long as it had (and, in particular, for longer than any other super-injunction). But that seems to me an unfair characterisation of Chamberlain J’s careful analysis in his third judgment. Moreover, in circumstances such as these, the passage of time is surely a highly pertinent consideration, bearing in mind that every passing week is an additional week during which Government decision-making on highly contentious matters involving the expenditure of extraordinary amounts of public money remain shielded from any public or political scrutiny whatsoever. In such a scenario, it is hard to escape the conclusion that the justificatory burden under which the Government finds itself inevitably grows heavier with the effluxion of time.
In any event, by the time of Chamberlain J’s fourth judgment this summer, matters had unambiguously moved on, because a Government-commissioned review had concluded that acquisition of the dataset by the Taliban was ‘unlikely to substantially change an individual’s existing exposure given the volume of data already available’ and that ‘it appears unlikely that merely being on the dataset would be grounds for targeting’. At paragraph 33 of his fourth judgment, Chamberlain observed that the assessments in the Government-commissioned review ‘are very different from those on which the super-injunction was sought and granted’ and that that could only partly be attributed to the passage of time. This raises the question whether, in the first place, there was ever actually a sufficient justification for granting the super-injunction. On this point, Chamberlain J concluded by noting that: ‘It will be for others to consider whether lessons can be learned from the way the initial assessments in this case were prepared and whether the courts were, or are generally, right to accord such weight to assessments of this kind.’ The latter is a question upon which it might be particularly fruitful for the Court of Appeal to reflect, the judgment of which caused the super-injunction to remain in place for almost exactly year more than it otherwise would have done.
Parliamentary accountability as a legal-constitutional principle
A theme that featured prominently in Chamberlain J’s successive judgments, but which appeared to commend itself far less to the Court of Appeal, concerned the importance of parliamentary scrutiny of the Government and of ministerial accountability to Parliament. In his first judgment, at paragraph 16, he observed that the super-injunction could not, thanks to parliamentary privilege, constrain proceedings in Parliament — it would not, for instance, be unlawful for an MP or peer to ask a question of a Minister revealing the existence of the injunction — but that in practice the role of Parliament was nevertheless effectively neutralised by the injunction:
Under Article IX of the Bill of Rights, no such constraint would be constitutional or lawful. I varied the injunction to make this point clear for the avoidance of doubt … Nonetheless, MPs and peers cannot ask questions about something they do not know about; and the Parliamentary authorities may regard the existence of an injunction a relevant to their decisions about what can and cannot be raised.
It follows that the practical effect of the super-injunction was to prevent Parliament from carrying out its imperative constitutional function of holding Minsters to account in respect of an issue that raised sensitive, complex and costly questions of public policy. Albeit in very different circumstances, the Supreme Court, in the Miller II case, was presented with a situation in which, thanks to being prorogued for several weeks without good reason, Parliament was similarly prevented from carrying out its constitutional functions. In that context, the Supreme Court was moved to observe (at paragraph 46 of its judgment) that the principle of parliamentary accountability — meaning Parliament’s holding of Ministers to account — had been
described by Lord Carnwath in his judgment in the first Miller case as no less fundamental to our constitution than Parliamentary sovereignty (R (Miller) v Secretary of State for Exiting the European Union [2017] UKSC 5; [2018] AC 61, para 249). As Lord Bingham of Cornhill said in the case of Bobb v Manning [2006] UKPC 22, para 13, “the conduct of government by a Prime Minister and Cabinet collectively responsible and accountable to Parliament lies at the heart of Westminster democracy”. Ministers are accountable to Parliament through such mechanisms as their duty to answer Parliamentary questions and to appear before Parliamentary committees, and through Parliamentary scrutiny of the delegated legislation which ministers make. By these means, the policies of the executive are subjected to consideration by the representatives of the electorate, the executive is required to report, explain and defend its actions, and citizens are protected from the arbitrary exercise of executive power.
So important did the Supreme Court consider parliamentary accountability to be that it made it clear in Miller II that the principle is no mere constitutional convention: it is also a legal principle that cannot lawfully be impeded or frustrated by the government absent reasonable justification. As noted above, the Afghan super-injunction litigation, unlike Miller II, was not concerned with the lawfulness of a purported exercise of power by the Government. But if, following Miller II, parliamentary accountability is a legal-constitutional principle with real bite, it is nevertheless surprising that it did not — in that form — warrant mention in any of the several judgments rendered in respect of the Afghan super-injunction. (A similar point is made by Paolo Sandro in a helpful thread on Bluesky.) At the very least, the weight ascribed to the principle of parliamentary accountability by the Supreme Court in Miller II serves to underline the correctness of Chamberlain J’s emphasis on examining the impact of the super-injunction upon the possibility of parliamentary scrutiny — which, in turn, demonstrates the appropriateness of close judicial scrutiny of the Government’s attempt to justify injunctive relief that effectively rendered parliamentary accountability a dead letter in this context.
The legal and constitutional importance ascribed to parliamentary accountability by Miller II also places in a questionable light a significant aspect of the Court of Appeal’s judgment in the super-injunction case. At paragraphs 63–65 of its judgment, the Court of Appeal held that Chamberlain J had, in his third judgment, fallen into error partly because of the weight he had accorded to the possibility, were the super-injunction to be discharged, of those Afghans named in the dataset but not slated for relocation to attempt to leverage public and parliamentary opinion:
The Judge, however, failed to follow through the logical sequence of events which would have to occur if such public and Parliamentary pressure were to be brought to bear. First, the injunction would have to be discharged – indeed, that is one reason why the Judge concluded that it should be discharged. But in reality, the very act of discharging the injunction would, on the undisputed evidence, immediately set in motion a chain of events leading to the dreadful (and potentially lethal) consequences which the Judge was seeking to avoid. The reality is that, very quickly after becoming aware of the data breach, the dangers and difficulties for people in Cohort C seeking to leave Afghanistan would be increased, even before the Taliban had obtained the dataset.
It is also important to note that all this would happen after the injunction was discharged. The Parliamentary pressure to which the Judge referred could not take place before discharge, for example during the grace period of 21 days which the Judge ordered. Although steps could be taken to help Cohort C in that grace period, for example to notify them of the data breach so that they could consider taking steps to help themselves, the wider publicity which the Judge contemplated would take place and which might lead to Parliamentary pressure on the Government could only take place after the end of the grace period.
If one stands back and considers the reality of what would happen in that scenario, it is obvious that there would be insufficient time for there to be the sort of Parliamentary pressure that the Judge had in mind. In other words, the discharge of the injunction would have precisely the opposite effect to the one intended by the Judge. We accept the submission made by Sir James Eadie KC on behalf of the MoD that the means chosen by the Judge had no rational connection with the aim he was trying to achieve; indeed, it would be self-defeating.
There is, of course, room for reasonable disagreement about whether Chamberlain J’s or the Court of Appeal’s overall evaluation of the likely consequences of the discharge of the injunction is to be preferred as matters stood a year ago. However, stepping back from the detail of that disagreement, it striking that the Court of Appeal’s characterisation of the value of parliamentary scrutiny is a highly instrumental one. On that court’s analysis, whether the relevance of parliamentary scrutiny appears to turn upon its capacity (or otherwise) to yield practical results. But that reflects a highly impoverished view of the principle of parliamentary accountability, the importance of which — as a fundamental axiom of constitutional democracy — transcends the capacity of scrutiny to generate particular outcomes.
That is perhaps a fitting point with which to conclude, for it serves to encapsulate the tension at the heart of this case, between the potentially horrifying human consequences of permitting the data breach to become known and the implications — which are more diffuse, but which are no less important for that — of permitting the Government, for a period of nearly two years, to enact policy, spend vast sums of public money and manage the consequences of catastrophic administrative failure behind an impenetrable veil of secrecy, shielded from all and any public scrutiny and political accountability. None of this implies that there are no circumstances whatever in which a super-injunction of this nature ought to be granted in favour of the Government — but it is surely a step that ought to be taken only in the most extraordinary circumstances, and a remedy that ought to be kept in place for the shortest time possible. Against that background, the decision made by Robin Knowles J on 1 September 2023 looks a lot more defensible than the decision, overruling Chamberlain J’s third judgment, issued by the Court of Appeal almost a year later.
Public Law for Everyone 