IT’s been a busy few weeks for surveillance and its practitioners. For the first time, a director-general of ASIS, Australia’s foreign intelligence agency, spoke publicly about the work of that organisation. A Canadian spy was accused of leaking Australian and US secrets to Russia. And the Gillard government has initiated a parliamentary inquiry into proposed wide-ranging reforms of Australia’s national security laws.
Commentators quickly labelled two of the government’s proposals as symptoms of a Big Brother state. The first would extend existing surveillance power to new technologies such as Twitter and Facebook; the second would require telecommunications providers to retain data about their customers’ activities for at least two years. One person was so incensed by the second of these proposals that he hacked into an Australian internet service provider, AAPT, and extracted information in order to show how “a lack of security at ISPs and telcos” might put Australians’ private information at risk. These are undoubtedly controversial proposals, but they are only two of more than forty set out in the inquiry’s terms of reference – and they are not the most worrying.
To begin, though, it is worth giving credit where it is due. A major problem with national security legislation is that it is often a knee-jerk response to a crisis, pushed through parliament without adequate debate about workability or the implications for civil liberties. The current inquiry is being conducted – as much as this can ever be said where national security is concerned – at a time of relative calm.
In the inquiry’s terms of reference, the government sets out three tiers of proposals: those it wishes to proceed with, those it is considering proceeding with, and those on which it simply wants the committee’s advice. The government has not prepared detailed legislation; instead, to guide the parliamentary inquiry and public submissions, it has released a discussion paper explaining some of the broad issues.
This is a good way of getting feedback on legislative proposals at a very preliminary stage, and is a significant improvement on the practices of the Howard government. In November 2005, that government recalled parliament to pass significant amendments to Australia’s anti-terrorism laws with just one day’s notice. In contrast, the main anti-terrorism law put forward by the current government, the National Security Legislation Amendment Act 2010, was the product of several years of consultation. First, the government released responses to six preceding inquiries into Australia’s anti-terrorism laws. Then the Attorney-General’s Department released a discussion paper for public consultation. Once prepared, the legislation itself was the subject of a further review by the Senate Legal and Constitutional Affairs Committee.
But although the government should be commended for abandoning the high-handed approach of its predecessor, to some extent the current terms of reference and discussion paper work against effective public consultation. The discussion paper doesn’t address each of the proposals in the terms of reference clearly and logically. In fact, several of the proposals – including the data retention plan, which is likely to be among the most controversial issues at the parliamentary inquiry – aren’t mentioned at all.
Even those proposals that are covered in the paper are not explained in sufficient detail, which makes evaluating them very difficult. In particular, the discussion paper neglects to specify the accountability measures that would attach to particular proposals. It simply says that the proposals will be “coupled with the appropriate checks and balances and oversight mechanisms society rightly demands” (our emphasis), whatever they may be. With the cumulative effect of the proposals representing a significant expansion in the powers of law enforcement and intelligence agencies to spy on Australians, it seems only common sense that accountability measures should also be expanded.
Yet the discussion paper implies at several points that existing accountability measures are outdated – that they “reflect historical concerns about corruption and misuse of covert powers.” Far from strengthening these measures, the government seems to be suggesting that they should be weakened. This is disturbing. Concerns about abuses of power are not merely “historical”: in 2004, for example, Justice Adams of the New South Wales Supreme Court said that ASIO had perpetrated a “gross breach of the powers given to the officers under the warrant” when it forced twenty-one-year-old terrorist suspect, Izhar Ul-Haque, to answer its questions. When Mamdouh Habib claimed that he had been tortured in Egypt in the presence of an Australian official, the inspector-general of intelligence and security, while rejecting the specific allegation, criticised Australian officials for not doing enough to prevent his mistreatment.
Apart from the lack of detail, the overwhelming problem with the discussion paper is that it fails to give enough attention to civil liberties in its quest to streamline and simplify the powers of law enforcement and intelligence agencies. It complains in great detail about the complexity of the telecommunications interception regime, highlighting the different criteria for law enforcement agencies to intercept a person’s real-time communications, such as phone calls, and to gain access to a person’s stored communications, such as cached material on a computer. But in proposing to standardise these criteria, the paper adopts a lowest-common-denominator approach. As a general rule, a person’s communications may only be intercepted to investigate an offence carrying a penalty of at least seven years’ imprisonment. In other words, privacy may only be invaded to investigate a very serious offence. But the government proposes to adopt the lower criterion that applies to stored communications – three years’ imprisonment. This would radically widen the circumstances in which a person’s privacy could be invaded. And not only the privacy of those people who have committed, or are suspected of, criminal acts: law enforcement agencies may also apply to intercept the communications of a suspect’s friends, family members, work mates or even spiritual advisers.
Another proposal in the same vein is to create a single category of warrant under the Telecommunications (Interception and Access) Act 1979. At present, there are two main categories of warrant. The first allows interception of any communications over a single service, such as a particular telephone number. The other covers interception of any communications passing over the services or devices belonging to a particular person – a far broader and more invasive option. Decision-makers will therefore only give law enforcement officers a named person warrant as a matter of last resort, for example, where it is impossible to identify the telephone number a suspect is using. The danger in combining the two categories of warrant is that law enforcement agencies will have a carte blanche to use any powers that they (rather than the decision-maker) regard as appropriate. And the tendency is for agencies to use all the powers that are available to them.
THE other major set of proposals deals with “equipping and enhancing the operational capabilities” of Australia’s intelligence agencies. These proposals, which involve a very considerable expansion of these agencies’ powers to spy on Australians, ring the loudest alarm bells for us. Because intelligence agencies operate without the same level of oversight and accountability as their law enforcement colleagues, any expansion of intelligence gathering powers should be treated with great caution. “A key element of risk management is our ability to remain secret and to operate in secrecy,” ASIS’s director-general, Nick Warner, said in his recent speech – but the consequence is a substantially increased potential for abuses of power.
Starting with ASIO, Australia’s domestic intelligence agency, the discussion paper outlines a proposal to merge all warrant powers into a single category. While this is similar to the proposal discussed above, it is even more worrying because of the larger number of powers available to ASIO. We are not simply talking about interception of communications: those powers also include searching private premises, installing listening devices, inspecting postal articles and using tracking devices.
The paper also proposes an expanded definition of “computer” for the purposes of an ASIO computer access warrant. Rather than covering a specific computer, this warrant could extend to any computers on a “particular premises” or even a “particular computer network.” Like many of the other proposals in the discussion paper, this would have a disproportionate impact on the privacy of innocent third parties. Let’s say, for example, that ASIO wanted to gain access to the computer of a University of New South Wales academic. The proposed amendment would mean that ASIO could access the thousands of computers on the university’s network, even if there was no evidence that the academic had used any computer other than his or her own. Other proposals would double the duration of an ASIO premises search warrant (from ninety days to six months) and create a new category of warrant allowing ASIO to search a person.
It may be that some – or possibly even all – of the proposals in the discussion paper are justified. But the government needs to demonstrate conclusively why this is so. The closest the discussion paper comes to a convincing justification is to say, for example, that “there have been instances where ASIO was unable to execute a search warrant within the ninety day limit for reasons beyond its control…” (our emphasis). Unless such instances are frequent, surely it is not too much of an administrative burden on ASIO to apply for another warrant?
Similar criticisms might be levelled against the final proposal relating to ASIO, which would give the agency’s officers immunity from civil and criminal liability when they are acting under cover. Neither Britain nor Canada nor New Zealand has pursued this option, and it is difficult to see how such an unprecedented power could be justified here. Law enforcement officers are already given broad powers to conduct undercover operations. Even where an ASIO officer does engage in unlawful conduct, the Commonwealth Director of Public Prosecutions can elect not to prosecute. The discussion paper gives the example of an ASIO officer who, in infiltrating a terrorist group, commits the offence of training with a terrorist organisation. But this example says more about the excessive breadth of Australia’s terrorism offences than it does about the need for ASIO officers to be given immunity from civil and criminal liability. The training offence, as with many of the terrorism offences, captures conduct well in advance of the commission of a terrorist act, and doesn’t distinguish between the types of training involved or the purpose of that training. Perhaps it is these offences – and not the position of ASIO officers – that are in need of reform.
The discussion paper’s proposed reforms are not limited to ASIO. They would also extend the power of the less familiar ASIS and of the Defence Signals Directorate, or DSD, and the Defence Imagery and Geospatial Organisation, or DIGO. These are, in essence, overseas spy agencies. To protect against abuses of power, the collection of intelligence about Australians has generally been left to the more (though still insufficiently) accountable ASIO. Under the proposals, there would be a significant change to this arrangement. A new power would be created for ASIS, DSD and DIGO to collect intelligence about Australians whenever the agencies are cooperating with ASIO in the performance of its functions. The only safeguard would be that cooperation must be requested by the director-general of ASIO and the minister must consent.
This parliamentary inquiry deals with very significant issues, but the discussion paper’s arguments for change are sketchy at best. The inquiry should therefore be regarded as part of an ongoing process of scrutiny – right up until any legislation is passed – rather than the last word. We must make sure that adequate accountability measures are included as part of any reforms. Simplification may well be a virtue but the rights of Australians must also be protected. The government has opened the debate, but it is vital that civil liberties are front and centre.
Jennifer Goh is Social Justice Intern and Nicola McGarrity a Lecturer in the Gilbert + Tobin Centre of Public Law at the University of New South Wales.
We thank Jessie Blackbourn, Fergal Davis, Keiran Hardy and George Williams for their input.