THE control orders regime under the Prevention of Terrorism Act 2005 – whereby potentially draconian restrictions upon freedom of movement, expression and association can be placed upon suspected terrorists simply on the basis of reasonable suspicion – is a highly controversial and much litigated one. An important milestone was reached in Home Secretary v. AF (No. 3) [2009] UKHL 28, [2010] 2 A.C. 269 (noted [2009] C.L.J. 495), in which the House of Lords held that Article 6 ECHR requires revelation to the suspect of at least the gist of the putative grounds for suspicion. Subsequently in AN v. Home Secretary [2010] EWCA Civ 869 the Court of Appeal was concerned with the remedial and practical consequences which arose when control orders were imposed, prior to the House of Lords' ruling in AF (No. 3), in breach of the disclosure obligation spelled out in that case.

Section 3(12) of the 2005 Act says that if a control order is “flawed”, then the court may, inter alia, quash it (retrospectively) or revoke it (prospectively). However, while this implies that the court may make a choice, whether both of these options are normally available turns upon the prior question whether the control order is void or voidable. If it is void, then it has never existed in law, making prospective revocation, which necessarily implies the legal existence of the measure up until the date of revocation, logically impossible. If, however, the unlawful order is merely voidable, the position is different, since this analysis presupposes the legal existence of the measure until and unless a remedy is issued. It follows that the ostensibly abstract distinction between void and voidable administrative action can have profoundly important practical consequences. Indeed, this was the case in AN itself. The appellant was charged with breaching his control order. If it was void, he could not be guilty of the offence, because there would never have been a legally extant order capable of breach (a view recently underscored by Ahmed v. HM Treasury (No. 2) [2010] UKSC 5, [2010] 2 A.C. 534, in which Lord Phillips acknowledged that the exercise of remedial discretion in relation to void administrative orders merely serves to obfuscate the logical implications of voidness). If, however, the control order was voidable, and prospectively revoked, it would have been in force at the time of the alleged infraction.

At first instance ([2009] EWHC 1966 (Admin)), Mitting J. had been influenced by certain peculiar features of the 2005 Act whereby a control order (i) cannot be made without the prior permission or (in urgent cases) subsequent ratification of the High Court, and (ii) is automatically subjected to more intense judicial review at a subsequent hearing. Mitting J. considered that this regime gave the Secretary of State's order a form of judicial imprimatur by virtue of court approval at stage (i), and that the Minister's procedural default did not crystallise until the stage (ii) proceedings, in which disclosure of the gist of the allegations against AN was required by AF (No. 3). This led Mitting J. to the conclusion that the control order was “lawful at inception” but could “no longer be sustained”, and that it should be prospectively revoked, not quashed.

In the Court of Appeal, Maurice Kay L.J. (with whom Rix and Stanley Burnton L.JJ. agreed) rejected Mitting J'.s analysis, preferring that of Silber J. in AE v. Home Secretary [2010] EWHC 42 (Admin) (which itself was the subject of an appeal by the Home Secretary in the instant proceedings) and following dicta in Home Secretary v. JJ [2007] UKHL 45, [2008] 1 A.C. 385 and MB v. Home Secretary [2007] UKHL 46, [2008] 1 A.C. 440. Maurice Kay L.J. was at pains to emphasise that “the making of a control order is an administrative act of the Secretary of State”, a status which is unaffected by the High Court's involvement at stage (i): control orders are not, therefore, court orders which for that reason (following Isaacs v. Robertson [1985] 1 A.C. 97) remain valid until set aside. The Court of Appeal was also unpersuaded that invalidity only arises at stage (ii), when the disclosure obligation identified in AF (No. 3) arises.

The Court of Appeal concluded that the orders in the instant cases (in which the Home Secretary had not disclosed even the gist of the case against the controlee) should be quashed ab initio, but the full extent of the legal foundations underpinning this conclusion are not entirely clear from the judgment in AF. Three possibilities arise. First, an Article 6(1) disclosure duty may arise when the order is initially made; non-disclosure at stage (ii) would then deprive those legal proceedings of their potential capacity to “cure” the earlier defect. However, given that the Act provides for compulsory judicial review, it is unlikely that the initial decision is determinative of civil rights and obligations, meaning that Article 6(1) is triggered only later. Second, as Maurice Kay L.J. pointed out, any executive assertion of the initial validity of a control order would turn upon the reasonableness of the Minister's suspicion, the grounds for which cannot fairly be tested precisely because of the refusal to disclose even the gist of the case to the person concerned. Unless the controlee makes no Article 6(1) objection at stage (ii), the court is therefore entitled to assume that the order was never valid. Third, there may be an implied statutory provision making it unlawful to exercise the order-making power if the Minister never intends to make adequate disclosure. If such orders were merely voidable, then the executive would, in practice, be able to impose control orders that would be temporarily valid even though they would necessarily be found unlawful at stage (ii). Maurice Kay L.J. was surely right to hold that in such circumstances – which appeared to arise in the instant cases, albeit that the Minister had not appreciated when making the orders (because AF (No. 3) had not been decided) how extensive the disclosure requirement would be – it should be unlawful “for the Secretary of State to begin to move towards the making of a control order”.

The point remains, however, that circumstances may arise in which an order imposed at stage (i) may be legally effective unless and until it is revoked (e.g. due to inadequate disclosure) at stage (ii). For example, the Home Secretary may impose the order at stage (i) intending subsequently to disclose the gist of the allegations; a later change of heart resulting in non-disclosure at stage (ii) would not engage any implied stage (i) prohibition upon imposing orders whilst intending not to disclose. In such circumstances, the administrative stage of the process would not be infected by any relevant illegality: the order would, as Mitting J. put it, be “lawful at inception” but would become unsustainable upon a stage (ii) refusal to disclose. This analysis does not call into question the general principle that unlawful administrative acts are void: it merely underlines the fact that Parliament may prescribe a decision-making structure in which certain legal obligations (such as the requirement to disclose the gist of allegations to controlees) crystallise only after the commission of the initial act, thereby preventing the breach of such obligations from depriving the relevant act of ab initio validity.