Freedom of Information Act 1982

Re Milliss and Australian Archives (Unreported, AAT, 23/8/96)

Department of Health & Anor v Jephcott [1985] FCA 370; (1985) 62 ALR 421

Searle Australia Pty Ltd v Public Interest Advocacy Centre & Anor (1992) 108 ALR 163

Re McKnight and Australian Archives [1992] AATA 225; (1992) 28 ALD 95

SYDNEY, 9 April and 25 September 1997 (hearing), 10 October 1997 (decision)

Solicitor for Respondent: Australian Government Solicitor

1. This is an application to review a decision made by the respondent pursuant to section 43 of the Archives Act 1983. The applicant made an application pursuant to section 40 for access to a range of documents held by the respondent after transmission of the documents by the Australian Security Intelligence Organisation (ASIO). The applicant essentially sought access to his own file. By reason of the age of the documents they fall within the open access period (s 3(7)) and are to be made available under section 31. Nevertheless, they are subject to claims for exemption under section 33 on the grounds set out in that section.

2. One hundred and twenty five pages were identified as falling within the terms of the request. Following consultation and conferences within this Tribunal and following a visit made by Mr Milliss to the respondent in Canberra at its request, some 105 pages were made available unreservedly to the applicant. Of the remaining 20 pages, claims for exemption were made in respect of 17 totally and in respect of 3 partially.

3. The applicant had previously made a request for access to documents relating to his father, Bruce Milliss. Following claims for exemption in respect of certain of those documents, two previous applications were made to this Tribunal to review the respondent's decisions. The decision of this Tribunal on both applications was given on 23 August 1996 (Decision number 11192). As many of the issues canvassed and decided in that matter have relevance in the present proceedings, we will refer to this decision from time to time as the first application. We agree with and adopt the statements of law set out in the reasons for that decision.

4. As in the first application, claims for exemption were based upon two paragraphs of section 33. Of the 20 documents, some derived from what the respondent referred to as `human sources'. The remainder derived from what were called `foreign liaisons'. In relation to all the documents, a claim for exemption was made under paragraph 33(1)(e)(ii) This provides for exemption if the record contains-

`[I]nformation or matter the disclosure of which under this Act would, or could reasonably be expected to . . . disclose, or enable a person to ascertain, the existence or identity of a confidential source of information in relation to the enforcement or administration of the law.'

5. Claims for exemption of records derived from foreign liaisons were principally made under paragraph 33(1)(a). This provides that a record is exempt if it contains information or matter the disclosure of which could reasonably be expected to cause damage to the security of the Commonwealth. The basis of the claim was that disclosure of the information in the documents or even in some cases, of the existence of a foreign liaison, could result in termination of a continued association with foreign intelligence sources which would rebound to the detriment of the Commonwealth. Having regard to the conclusions to which we have come, it will not be necessary to consider the claim for exemption made under this paragraph.

6. To sustain an objection under paragraph 33(1)(e)(ii), all that is necessary is to demonstrate that the disclosure of information or matter in the document could reasonably be expected to disclose, or enable a person to ascertain, the existence or identity of a confidential source of information in relation to the enforcement or administration of the law. For reasons set out in paragraphs 11 and 12 of the first application, it is clear that the information contained in all the documents related to the administration of the law under which ASIO and its predecessors operated. For reasons set out in paragraph 6 of the first application, it is also clear that there is no element of public interest to be balanced or taken into account in considering the application of this paragraph of the Act.

7. In relation to a corresponding provision in the Freedom of Information Act 1982, Forster J said in Department of Health & Anor v Jephcott [1985] FCA 370; (1985) 62 ALR 421 at 425-

`However, there seems to be with all respect to the Tribunal to be no warrant in the Act or elsewhere for engaging in the exercise of balancing one public interest against another and concluding that `substantial risk' must be shown before the identity of a confidential source whether actual or hypothetical is entitled to protection.

Section 37(1) makes no reference to public interest or competing public interests as do some other sections of the Act: see section 36 for instance which, or course, immediately precedes section 37.

If a source of information is properly to be regarded as a confidential source then there is, in my view, no room for the concept of `a substantial risk that the administration of the law will be impaired'. Either it is established that the source is a confidential source or it is not. If it is, then it is entitled to the protection given by section 37(1)(b) and no `bearing in mind' of the purposes of the Freedom of Information Act either as set out in section 3 or discovered elsewhere can, in my view, affect the position.'

8. All of the foreign liaison documents disclose the identity of the originating organisation. This is done either explicitly through letterheads, signatures, references and the like, or implicitly, through layout, form and recognisable correspondence habits. All of the foreign liaisons, whose documents are included in the 20 core documents the subject of this application, have been approached and have confirmed that the information contained in the documents was supplied in confidence on the basis that the confidence was intended to continue indefinitely. Even without this evidence it would have been clear from a simple examination of the documents themselves, that the source of the information contained in the documents was a confidential source.

9. The material in the human sources documents is capable of disclosing the identity of the relevant sources of information. All of them must be regarded as confidential sources. Again, the nature of the documents and the manner in which they came into existence indicate, without the necessity of extrinsic evidence, the confidential nature of the human sources. Sometimes their identity is explicitly stated. At other times it can be concluded by cross-matching of information in documents known to be in the possession of the applicant (through archive access) or in documents that are readily available, such as documents held at the Mitchell Library. In all cases, the confidentiality must be presumed to have continued. Evidence was given similar to that referred to in paragraphs 24 and 25 of the first application. Reference was also made to a method of identifying sources through the application of the logical process of induction. This is sometimes referred to as the mosaic analysis method. In paragraph 8 in the reasons for decision for the first application, reference is made to the previous decisions of this Tribunal, where such a method of analysis has been accepted as a valid method of approach in determining whether identities could reasonably be expected to be disclosed in documents of this nature.

10. The applicant vigorously urged us to reject the mosaic analysis method. In a flow of language he referred to it as "essentially empty, sophistic, fundamentally fallacious, false, misleading, capricious, arbitrary, fickle and exalted humbug". He submitted that it was a doctrine invoked by ASIO only when it did not wish to release documents.

11. It was a little difficult for us to understand the logic of the applicant's approach. He argued that if, as the evidence showed, some documents required only 1 or 2 pieces of information to enable a source to be identified then, "this is not much of a mosaic". In his view, a proper mosaic consisted of hundreds, if not thousands, of pieces. This is confusion based upon a word used figuratively, for which many other words could be substituted. In any event, one would have thought that a document that gave up its secrets to a key of only one or two pieces of information should be much more closely guarded than one that required hundreds or thousands of keys. A document with few gaps could more reasonably be expected to disclose the identity of a confidential source than one lacking multifarious pieces of information. It may be, however, that the applicant was not so much concerned with the effect of likely disclosure, as he was with the insistence of ASIO upon explaining its claims with reference to a process which Mr Milliss thought fallacious.

12. However, whatever name is given to the process by which comparison of one or more pieces of information could lead to the identification of a source, the fact is that the evidence supports the possibility of such a process and the use by those concerned to discover identities through such methods of approach. Having regard to the actual material before us, we have concluded that such a discovery is not a mere possibility but could reasonably be expected if ordinary techniques of intelligence analysis are applied to the documents. The possibility of those techniques are, as Mr Milliss has pointed out, invoked with some regularity by ASIO in claims for exemption under this paragraph. That is not the only purpose they serve however. The same techniques are regularly used by all intelligence services for their own purposes, not only in supporting claims for exemption.

13. As in the first application, the applicant indicated that he had no desire to have access to names, signatures, initials of ASIO officers, ASIO file numbers and the organisation structure of ASIO and ASIO codes and codewords. All of these appear in various parts of the human source documents. He also said that he did not wish to know the identity of these sources. Nevertheless, he admitted with commendable candour (transcript page 17)-

`I would like to reiterate that I am not interested or concerned in learning the identity of confidential sources whatsoever. I would be less than human, though, members of the Tribunal, if I said that I would not have natural curiosity about the identity of those sources.'

14. Whether or not the applicant could restrain his `natural curiosity', the fact is that release of these documents would be release to the world. As was said in Searle Australia Pty Ltd v Public Interest Advocacy Centre & Anor (1992) 108 ALR 163 at 179- `Disclosure under the FOI Act is, of course, disclosure to the public, and the particulars and personality of the applicant are of no significance.' Because of this, there is no possibility of control over the use to which any of the subject documents could be put. There is no way in which this Tribunal or any agency could accept any undertaking not to use the documents adversely to the interests of any of the sources. Whether the applicant declined to undertake the deductive processes involved in identifying the sources, there is no way in which any other person could be so restrained.

15. The applicant invited us to depart from previous Tribunal decisions in which the so-called mosaic theory was explained and adopted. An example is in Re McKnight and Australian Archives [1992] AATA 225; (1992) 28 ALD 95 at 105. Acceptance of factual evidence in one application or, indeed, similar evidence in many applications does not amount to the evolution of rule of law. In accepting evidence put before us in the present application that sources are identifiable in the human source documents, through internal or external analysis, and with or without the application of other known facts, we are doing no more than accepting the evidence put before us in this application.

16. We are, however, prepared to say that such a process is better described as an induction process, if for some reason the use of the word `mosaic' troubles any observer. There is no special reason why this well-known process of logic should be described figuratively. Induction is defined relevantly in the Macquarie Dictionary as `the process of discovering explanations for a set of particular facts, by estimating the weight of observational evidence in favour of a proposition which asserts something about that entire class of facts' and secondly as, `a conclusion reached by this process.' What may be done in analysing the subject documents, and the particular facts recorded in those documents, is to estimate the weight of internal evidence or external connecting evidence in favour of a proposition which is likely to disclose the identity of a confidential source. The so-called mosaic theory is nothing more than this.

17. In a previous application, the applicant was successful in obtaining two documents which were identical in every respect except that in one, the name of a person had been expunged, and the in other the name was left for all to see. The applicant considered that by releasing the earlier document including a person's name, ASIO had "established a new benchmark". Having revealed the name of a person on one occasion, the applicant's submission, as we understand it, was that thereafter there was no basis upon which ASIO could refuse to release other names of other persons in the future. In relation to that submission, we make two observations. Firstly, the document (of which there were two copies) is a record of interview between the person named and an ASIO officer and deals with the named person's application for a visa to enter the United States of America. It is totally different in its nature from the human source documents, the subject of our present consideration. It is a record of an interview sought by the named person with an officer of ASIO in order to clarify certain matters so as to facilitate his travel plans. The human source documents are of an entirely different nature. They identify, or are capable of identifying, a person providing information in relation to the administration of the law. The document on which the applicant relies provides no such information and could not be the subject of a claim for exemption under paragraph 33(1)(e)(ii). The second observation we would make is that even if there had been an inadvertent release of the name of a confidential source of information, that in itself could not justify the release of another document containing material which could identify another confidential source of information. There is no question of a "benchmark" having been set, so as to legitimise any further disclosure which might compound the original erroneous disclosure.

18. The applicant also invited us to speculate as to the likelihood of foreign liaison sources refusing to cooperate with ASIO in the future if any of the foreign liaison documents was disclosed. Imagine, he invited us, that a foreign liaison agency became aware of a plot to assassinate the Australian Prime Minister. Could we suppose, he asked, that they would fail to pass this information on because of a minor infringement of confidence that happened in the past. This argument relates to the claim for exemption under paragraph 33(1)(a) and, for reasons to which we have already referred, does not fall for consideration. Evidence which was received in confidence, and to which reference cannot be made in detail did, however, indicate that there has been at least one break in relationship between two otherwise allied foreign liaisons, brought about by a minor breach of confidence. However, in relation to the foreign liaison documents in the present application, all that is necessary to be shown is that the disclosure of the documents would disclose both the existence and the identity of the foreign liaisons as confidential sources of information. For reasons to which we have referred, we are of the view that this has clearly been made out.

19. Evidence given similar to that referred to in paragraphs 21 to 26 of the reasons for decision in the first application, leave us in no doubt that the elements of the exemption have been demonstrated. All of the foreign liaison sources have confirmed the nature of the original confidence under which the information was given. Notwithstanding the age of some of the information and, in some cases the trivial nature of the information, the fact is that the confidence invoked upon the giving of the information was such as to demand its respect indefinitely. In so far as it has been possible to confirm with surviving sources their understanding of the original degree of confidence, the same result obtains. Where it has not been possible to confirm positively with named persons their understanding of the initial relationship, the pattern of conduct disclosed and the documents themselves, confirm that we are entitled to infer a degree of confidentiality continuing indefinitely which can be released only upon the concurrence of the donor of the information. All of the documents sustain a claim for exemption under paragraph 33(1)(e)(ii).

20. That being so, it is not necessary to examine the claim for exemption based upon subsection 33(1)(a). One claim is sufficient to qualify all 20 documents as exempt records. The decision under review, therefore, is affirmed.