Australia’s integrity paradox: Criminal suspects may stay silent while public servants can be compelled to speak

Abstract

This Comment examines the widening gap between criminal procedure protections against compelled self-incrimination and the compulsory fact-finding powers exercised by Australian integrity and anti-corruption bodies. It argues that while such regimes are justified by different institutional purposes, coercive examinations and secrecy directions can create prolonged reputational and procedural harm if not governed by clear thresholds, time discipline and transparent safeguards. The work proposes tighter legislative and policy guardrails, including regular review of secrecy, improved public reporting and stronger procedural fairness architecture around compulsory questioning.

Australians are used to thinking of the criminal justice system as the place where the State exercises its harshest powers. That is true in many respects, but it obscures an uncomfortable legal reality. In ordinary criminal investigation, a suspect will usually be cautioned and told, in substance, that they do not have to answer police questions. By contrast, in many integrity and anti-corruption investigations across Australia, a witness, including a witness who is plainly under suspicion, can be compelled to attend, compelled to answer, compelled to produce documents, and exposed to criminal penalties for refusal. In many regimes, failure to attend, refusal to be sworn, refusal to answer a required question, failure to produce documents or even failure to volunteer information which is relevant (even if not directly in response to a question posed by an investigator) is itself an offence, sometimes punishable by imprisonment. For example, the National Anti-Corruption Commission Act 2022 (Cth) (NACC Act) creates offences for refusing or failing to answer required questions at a hearing. Beyond offence provisions, some Commissions have contempt mechanisms that can lead to court proceedings and custodial outcomes, including coercive detention logic in which custody can continue until the contempt is purged (that is, until compliance). The legal explanation is familiar and often sound, but the democratic and procedural consequences are still not being discussed seriously enough.

Criminal justice vs coercive examination

The orthodox distinction is easy to state. Criminal process is designed to adjudicate guilt and punish crime, so it gives strong protection against compelled self-incrimination. Integrity and anti-corruption bodies, by contrast, are designed to uncover facts, expose corruption, and protect institutions, especially in settings where ordinary investigative methods may fail because of closed networks, mutual protection and a culture of silence. That distinction matters, and any serious argument in this area must concede it at the outset. The issue is not whether integrity agencies should have coercive powers. They should. The issue is whether Australia has allowed coercive compulsion and secrecy in those regimes to outrun the safeguards, transparency and disciplined oversight needed to maintain public confidence.

Every Australian jurisdiction now has an integrity or anti-corruption body with coercive investigative powers, including the Commonwealth’s National Anti-Corruption Commission (NACC) and state and territory counterparts such as the New South Wales (NSW) Independent Commission Against Corruption (ICAC), Victoria’s Independent Broad-based Anti-corruption Commission (IBAC), Queensland’s Crime and Corruption Commission (CCC), and equivalents elsewhere. These bodies investigate alleged corrupt conduct and serious misconduct in public administration, including bribery, fraud, procurement and grants manipulation, abuse of office, undisclosed conflicts of interest, and improper influence over public decisions. Taken together, there are at least nine principal commissions across the federation, and the trend has been towards expansion and consolidation of coercive integrity powers, most visibly with the establishment of the NACC at the federal level.

The criminal law side of the comparison is familiar enough to most lawyers, but it is worth restating because it sharpens the contrast. At the Commonwealth level, the cautioning framework in the Crimes Act 1914 (Cth), including s 23F, reflects the basic principle that a person under official questioning is not obliged to say or do anything, even though anything said or done may be used in evidence. That is then reinforced, in Uniform Evidence Act jurisdictions, by provisions such as Evidence Act 1995 (Cth) s 89, which generally limits adverse use of pre-trial silence in official questioning. NSW’s Evidence Act 1995 departs from that baseline in a significant but carefully conditioned way through s 89A, which permits adverse inferences in certain serious indictable matters only if strict statutory preconditions are met, including a special caution and legal practitioner presence. In other words, even where the criminal law qualifies silence, it does so overtly and with procedural architecture that signals the exceptional nature of the departure.

The integrity and anti-corruption model works differently from the start. It is not built on a baseline of non-compulsion. It is built on statutory compulsion, then moderated by privilege and immunity rules whose scope depends entirely on the Act in question. At the federal level, for example, the NACC Act contains an express self-incrimination provision, s 113, which is part of a coercive examination framework rather than a police interview model. In NSW’s Independent Commission Against Corruption Act 1988 (NSW), especially ss 37 and 38, is a classic illustration of the same structure, witnesses are not simply free to refuse to answer on ordinary self-incrimination grounds, but the statute then addresses objections, declarations and later use. The ACT’s Integrity Commission Act 2018 (ACT) is another useful marker, because ss 175 and 176 make the point in a particularly explicit form by excluding common law self-incrimination privilege in the examination context and then providing statutory use and derivative use protection. Tasmania remains a notable point of contrast in parts of its scheme, with the Integrity Commission Act 2009 (Tas) s 92 preserving an express claims-of-privilege mechanism. The broad national pattern, however, is clear enough – compulsion first, then statutory qualification.

The ‘fairness problem’

That design may be legally coherent, but it produces a fairness problem that is too often minimised in public debate. A statutory ‘use immunity’ is not the same thing as a right to silence: direct ‘use immunity’ generally means the compelled answers cannot be tendered against the person; ‘derivative use immunity’ goes further and restricts use of evidence later obtained as a result of those compelled answers. Notably, many schemes provide direct use immunity but not derivative use immunity, which leaves real (ensuing) forensic and practical exposure even where the compelled testimony itself cannot be tendered. However, this is a downstream evidentiary protection. It may help limit direct admissibility in later criminal proceedings, and in some jurisdictions it may extend further through derivative use restrictions. But it does not remove the immediate burden of compulsion, and it does not answer the wider practical consequences of being forced into a coercive investigative process. For many people, especially public servants, local government officials, ministerial staff, contractors, or office holders, the immediate harms are not abstract legal questions about admissibility – they are reputational damage, career paralysis, legal costs, stress, loss of income, and the inability to explain publicly what is happening while an inquiry proceeds.

This is where the comparison with criminal process becomes politically and morally potent, even if lawyers instinctively resist it. The point is not that corruption witnesses are identical to criminal suspects. The point is that the State is using a severe form of coercive power, sometimes in private, against people who may not be charged with any criminal offence, while requiring cooperation on pain of prosecution. The public is entitled to ask whether the safeguards in that system are genuinely commensurate with the power being exercised. It is not enough to say, as a reflex, that integrity agencies are not criminal courts. Coercive State power does not become benign merely because it is classified as investigative or administrative.

There is a second and under-discussed problem that deserves much more attention, and that is time. Secrecy is not just cultural. It is frequently hardwired through statutory non-disclosure mechanisms, such as non-disclosure notations attached to compulsory notices and confidentiality directions attached to hearings, backed by criminal offences for breach. In the NACC scheme, for example, non-disclosure notations and confidentiality directions operate alongside offence provisions for contravention, which is why delay and administrative drift can translate into long-running, effectively gagging constraints. Even where a secrecy direction or non-disclosure requirement may be justified at the beginning of an investigation, the practical effect of these orders can become much more troubling when combined with long timeframes, administrative inertia and opaque extensions or continuations. A witness may remain under a silence notice, or under the practical shadow of one, for months or even years while the matter proceeds slowly, priorities shift, or decisions are deferred. In that period, the person’s capacity to defend their reputation, explain events, or seek support may be severely constrained, yet there is often little publicly visible scrutiny of how long such restrictions endure, why they remain necessary or whether they are being actively reviewed. Even if the legal framework technically allows for variation or revocation, the lived reality can look very close to an indefinite gag, particularly where review is opaque and renewal or continuation occurs without visible external scrutiny.

Accountability and the rule of law

That temporal issue matters because it exposes a broader accountability gap in the current public conversation. Much of the debate about integrity agencies focuses on the power and outcome any media headlines may generate, summonses issued, hearings held, reports released, referrals made. Far less attention is paid to process duration, the rate and duration of secrecy directions, the use of compulsory examinations in matters that do not lead to public findings, or prosecutions for non-cooperation and related contraventions. Without better public reporting on these operational features, we end up arguing in slogans. Supporters say coercive powers are necessary. Critics invoke ‘Star Chamber’ imagery. Support tends to come from integrity agencies and anti-corruption advocates who argue that closed networks and ‘code of silence’ dynamics cannot be penetrated without compulsion. Criticism tends to come from civil liberties perspectives and professional rule-of-law bodies, including the Law Council of Australia, which has repeatedly argued that where self-incrimination privilege is abrogated, stronger safeguards such as derivative use immunity and tighter secrecy controls should follow. Neither side is forced to confront the granular questions that should actually matter in a rule of law system: how often are these powers used, for how long, under what internal thresholds, with what judicial or other review, and with what consequences for people in the absence of adverse findings against them, and without any subsequent prosecution and conviction?

None of this is a case for weakening anti-corruption enforcement. It is a case for strengthening its legitimacy by tightening the link between coercion and accountability. At a minimum, legislatures and agencies should move towards clearer recorded thresholds for compulsory examinations, stronger and time-limited secrecy directions with reasons and review pathways, better plain-language explanations to compelled persons about what immunities do and do not protect, and more transparent public reporting on the frequency and duration of coercive orders. Legal representation should also be treated as central to legitimacy in these settings, not as a procedural afterthought. If the law is going to compel speech on pain of prosecution, the least it can do is ensure that the person being compelled understands the statutory framework governing that compulsion.

The deeper point engages rule-of-law and constitutional values, even if it does not necessarily raise a question of constitutional invalidity – all the more so in a country without a national bill of rights of the kind that some assume would solve these issues automatically. The more a legal regime departs from familiar protections, especially the ability to remain silent and the ability to speak publicly about what the State is doing to you, the more care should be taken with oversight, duration and transparency. That is not indulgence toward corruption. It is discipline about public power. A mature integrity system should be both effective and restrained, and it should be able to demonstrate that restraint rather than merely assert it.

Australia is right to invest in anti-corruption institutions. It is right to recognise that corruption can be hard to detect and harder to prove. It is right to arm integrity agencies with powers that ordinary investigators may lack. But it should also be candid about the paradox it has created. In one part of our legal system, a person suspected of serious crime may refuse to answer questions. In another, a public servant or other witness may be compelled to answer in secret, prosecuted for refusal, and constrained from speaking about the process for long periods, under orders that can be difficult to scrutinise from the outside. That arrangement may be legally rational in design, and sometimes operationally necessary, but it is not beyond criticism. In a democracy, criticism of coercive secrecy is not an obstacle to integrity – it is one of the conditions.