Top Secret Intelligence in Europe: A Tipping Point in Luxembourg? (Guest Post)

This guest post is by Deirdre Curtin, Professor of European Law at the University of Amsterdam and Director of the Amsterdam Centre for European Law and Governance (ACELG).  This term, she is also a visiting fellow at NYU Law School.  This post originally appeared originally on acelg.blogactiv.eu and we are pleased to cross-post it here.

Kadi is back in Luxembourg and with a vengeance! The timing is interesting both for the case itself and more generally for highlighting the use of secret intelligence and evidence to justify detention and other sanctions.

To start with the case itself, the oral hearing of the new Kadi case (Kadi 2) took place in Luxembourg earlier this week before the Court of Justice of the EU (CJEU). After 11 years on a UN freezing of assets list, Mr. Kadi was actually de-listed by the UN Al Qaeda Sanctions Committee some 10 days prior to that and then, a week later, by the European Commission. The immediate issue is of course one of the (continuing) admissibility of the appeal. Is there in fact still an interest in the CJEU adjudicating the case in these new and changed circumstances? The appealing Union institutions, the Commission and the Council, (supported by a total of 13 Member States intervening in the procedures) argued during the oral hearing that the interest is in having “the law” authoritatively clarified on the matter. For Mr Kadi rather the position is that the appeals should be dismissed as having no further legal significance since the Commission had withdrawn its own regulation 2 days before the hearing so it has no interest in continuing to argue for its legality.

Almost irrespective of what the Court actually does in this particular case, this long strung out litigation (together with the many other sanctions decisions challenged before the Luxembourg courts in recent years and still pending where the issue of secret evidence is more directly at issue) is of much wider structural significance in terms of the secrecy culture that is entrenched (and inter-locked) at all levels of government, including increasingly the supranational level. This structural issue is likely to continue to come before the Courts in Luxembourg in a variety of ways given the Union’s expanded practice in the fields of internal and external security (on the supranational context of this more general security landscape, see my 2011 lecture Top Secret Europe).

The core issue that I want to raise here is whether governments at all levels of governance (national, supranational and international) can continue to deny courts (and in particular the CJEU) access to highly classified information so that they can provide judicial review on the merits. Will the CJEU more generally and definitively establish the limits on EU executive power (Commission and Council but potentially also applying to other actors in the future, such as Europol or the European External Action Service) in terms of secrecy and refusal of access to classified information shared with other actors both nationally and internationally (and vice versa)? Or at least provide some more substantive guidance in this respect that relates to the underlying facts in the case before it?

To go back to the facts, Mr. Kadi was listed by the UN originally in 2001 on the basis of (presumably) ‘top secret’ classified intelligence provided by one or more members of the Security Council. But the problem seems to be actually more acute: as the UN Special Rapporteur on human rights while countering terrorism pointed out in 2010, the UN Sanctions Committee taking the listing decision may itself not ever have full access to the classified information held by the State seeking the listing. In other words, the practice of listing individuals by the UN may be serially taking place on the basis of secret and undisclosed intelligence.

In line with standard security practices and the originator control principle (ORCON), originators of intelligence material in any event retain control over whom this information is revealed to. This means in practice that such (highly) classified information will not be revealed to those listed or their lawyers but also not necessarily to the UN itself nor to regional organizations such as the EU who take their own (derivative) listing decisions.

When it comes to the EU implementing UN lists of terrorist suspects, it too does so without ever having fully seen or possessing all the evidence in question to support that decision with regard to named individuals. If it does not possess that evidence it is also unable to submit it to the Court in the event of a legal challenge and despite being ordered to do so by the Court. This is a structural problem for the Court.

When it comes to intelligence information supplied by EU Member States in the context of what are termed the autonomous EU lists (e.g. France in the earlier case involving the PMOI) the situation is different at least in theory. The Courts can demand to receive access to the classified information that led to the EU listing but here too –and despite the fact that the Council actually possesses the information in question- they will not necessarily receive it. This is because of the operation of the principle of originator control as a matter of EU rules as well as specific national rules.

In the PMOI case for example even though the Council had (most of) the actual information in its possession and it had in fact been distributed to the other Member States, France could – and did – refuse access to the Court. It could do so because of the application of the principle of originator control and the fact that its national law required it to do so in the circumstances.  The Council subsequently produced one ‘confidential’ document to the Court that was not provided to the PMOI (as well as other non-confidential documents) but could not produce the documents France refused. It is arguable that at the very least it is the Council itself that needs to assess whether the need for ‘confidentiality’ claimed by a Member State (France) is genuine and sufficient. Even then, it is up to the Court to review in full the exercise of the Member State discretion. It can only do so if it is given access to the confidential documents in question.

There is clearly a Catch 22 situation here: national and third party classification rules and systems may result in the courts simply not being given access. This constitutes the bulk of classified information being shared at the Union level, subject to the originator control principle. The easier question is whether the Council and other Union institutions and agencies can deny the Luxembourg Courts access to classified information that requires a security clearance (confidential or above) that has been actually classified at the supranational level (European Union Classified Information)? Here it seems fairly obvious that they cannot, even if they have used other sources in drawing up that document (the principle of derivative classification).

If the Court requests such EU documents/information as part of ongoing judicial proceedings then Union institutions (and agencies) cannot plead their own internal rules of procedure and decisions based on them to refuse access to the Court. In practice, the General Court is particularly active in, as a “measure of inquiry” (Article 65 of its Rules of Procedure), seeking and receiving access to EU classified information that is denied to the applicants and other parties and in keeping that information confidential from the other parties to the proceedings (a type of de facto closed material procedure).

Access to classified information is the key in terms of effective accountability also in legal terms by a judge or tribunal. The question is by whom and how. What worked ultimately in the UN system after 11 years in terms of results (de-listing) is the fact that under the new special procedure instituted by the UN in the aftermath of the first Kadi judgment the (newly created) Ombudsperson was able to negotiate (some) access to classified information in individual files. On this basis she (Judge Kimberly Prost) makes recommendations to the UN Committee in question based also it seems on access she has obtained to classified intelligence evidence by States supporting the original and continued listing.

Despite an ongoing struggle by the Ombudsperson to actually obtain access to classified information, a process that she has negotiated bit by bit and with some success with a limited number of individual states who are frequently an object of listing petitions (for the latest list of these States and types of agreements, see here), she has already a considerable track record of recommending and obtaining de-listing decisions. She produces bi-annual reports and the success rate is certainly impressive: in a total of 21 completed cases by the Ombudsperson (since mid-2010) only one request for de-listing was denied by the UN Sanctions Committee (one was withdrawn and one is still pending). In a total of 18 cases the decision by the UN Committee was to delist (individuals or entities; one list was ‘amended’).  Mr Kadi is the latest to be delisted to date.

Yet the issue of access to classified information remains absolutely critical and by no means a foregone conclusion in the UN context. In the words of the Ombudsperson herself, in her latest report to the Security Council: “the challenges related to access to classified/confidential information remain pressing and significant… further progress is urgently needed…lack of access to confidential/classified information has been a concern in at least four recent cases.”

As we have seen, the challenges related to access to classified/confidential information are equally omnipresent in the EU system where the CJEU has far-reaching judicial powers in individual cases brought before it and the responsibility to ensure effective judicial protection. The crucial question that the CJEU is going to have to sooner or later address is: what procedural system can it for the future devise at the level of the supranational judicial system to deal with the issue of classified material being viewed by persons with the appropriate level of security clearances (either members of the Court itself or security cleared ‘special friends’ of the Court as in some national systems)?

Or should the CJEU institute –formally- a closed material procedure? Article 79 of the new amended rules of procedure of the CJEU that will enter into force on 1 November 2012 enables cases to be held “in camera” for “serious reasons related, in particular to the security of the Member States…” and without publication of such oral proceedings. At the same time holding the oral part of legal proceedings “in camera” is not the same as laying down a specific procedure that enables (member of) a court (security cleared) or a type of ‘special advocate’ to access classified information and keep it confidential from the applicants but nonetheless act on their behalf. Yet, in Canada reliance on expert judges was found to be constitutionally insufficient and in the UK reliance on a regime of ‘special advocates’ has been challenged and narrowed by various requirements that demand the ‘gist’ of the allegations to be disclosed.

Luxembourg is being closely watched – not only in Brussels and national capitals but also in New York and perhaps closest of all – in Washington. At the hearing in Luxembourg in Kadi 2 this week, a number of lawyers from the US State department in the courtroom were present, observing the proceedings and listening keenly to the exchanges between the judges and the lawyers for the parties to the proceedings. They too understand that the case can have much wider implications than just for the future possibilities of Mr. Kadi to initiate further legal proceedings (seeking damages or otherwise).

Beyond secrecy, this case raises very squarely the issue of accountability of executive power in the composite legal and political orders of the EU. It shows the manner in which national executive power intertwines with supranational executive power and that in turn with international /global executive power and the risk that it may ultimately be ‘unbound’ in the absence of accountability forums who can adjudicate in individual cases on the basis of all the information (see further my book Executive Power in the European Union from 2009). The Court of Justice holds an important key in this sense and it will be revealing to see how it will handle classified information in the wider context of effective judicial protection.

What is however missing is an open public debate that reaches beyond the (governmental) insiders’ interest in preserving the existing system come what may (thirteen Member State governments lined up with the Commission and the Council against Mr. Kadi). There is--sadly--virtually no space for an open debate on what broadly constitutes the public interest. This is not necessarily consonant with the interest of the executive power itself at whatever government level it is operating.

There is of course in any democratic society a need for necessary secrecy on security related matters but it is not ‘unbound’: it must be limited and made accountable in one form or another. Time will tell whether a tipping point has been reached in Luxembourg at least with regard to the use of secret intelligence by the supranational executive power in a manner that deprives individuals of their rights to due process and effective judicial protection. The challenge is to find a procedural way of balancing claims that intelligence sources and methods must be kept secret in a less rights invasive manner than the use of secret evidence in non-criminal procedures. The Luxembourg courts, unlike the UN Ombudsperson and the UN Sanctions Committee, have the power to provide a judicial remedy to a person who has been listed, perhaps on the basis of flawed secret intelligence used as evidence to justify the listing. For now, all eyes will continue to watch Luxembourg.