Allegra Spender (Wentworth, Independent) Share this | Hansard source

I rise in support of the Secrecy Provisions Amendment (Repealing Offences) Bill 2026. This is some of the most significant reform to Commonwealth secrecy law in decades. It draws on two major reviews: the Attorney-General's Department Review of Secrecy Provisions completed in 2023 and the subsequent independent statutory review by the Independent National Security Legislation Monitor. I commend the government for acting on these findings.

Secrecy provisions are necessary. They protect national security and intelligence and ensure that those trusted with sensitive information can be held to that trust. But, as the Human Rights Law Centre put in their submission, 'Secrecy laws which are disproportionate and not properly calibrated are bad for Australian democracy and the rule of law.' This bill moves meaningfully closer to that balance, though not all the way there. I understand the bill has been referred to the Senate Legal and Constitutional Affairs Legislation Committee. I hope the issues I raise today are considered carefully and that the government responds to the committee's recommendations before passage.

The 2023 AGD review had confronting findings. Commonwealth legislation had accumulated around 875 secrecy provisions, which was enough for the New York Times to describe Australia as 'possibly the most secretive democracy in the world'. Of these 875 provisions, 168 were identified as no longer warranting criminal liability at all. The review instead deemed that civil or administrative sanctions were more proportionate. The AGD review also recommended a new general secrecy offence to replace the sunsetting section 122.4 of the Criminal Code, which had long been the mechanism through which the non-disclosure duties across Commonwealth legislation attracted criminal liability.

In 2024, the INSLM reviewed the general secrecy offences in part 5.6. It found them uncertain, inconsistent with rule-of-law principles and, in some cases, disproportionate, particularly as they applied to nonofficials such as journalists, lawyers and human rights advocates.

This bill gives legislative effect to both reviews. It repeals more than 300 secrecy provisions. This is nearly double the number recommended for repeal by the AGD review and a reduction of more than a third of Commonwealth secrecy provisions. This is achieved primarily through the repeal of section 122.4 and targeted removal of criminal liability from non-disclosure duties across 13 portfolio areas. I am pleased to see this.

The bill also introduces a new targeted secrecy offence directed at those who improperly use or communicate Commonwealth information to obtain a benefit or cause detriment. This narrower formulation responds to the PwC tax confidentiality breach but does consciously reject the broader general offence proposed by the 2023 AGD review.

In response to the INSLM's recommendation, the bill repeals the security classification limbs of the deemed harm offences, removing reliance on administrative classification markings as a criminal offence element. It repeals the 'dealing with' offence for nonofficials under section 122.4A(2), ending the position where journalists face criminal liability merely for receiving sensitive information. It narrows deemed harm offences for officials, requiring a more direct connection to actual or likely harm. It nearly halves the maximum penalties for non-official disclosure offences. It legislates a requirement that the Attorney-General's written consent be obtained before any journalist can be prosecuted for a secrecy offence. And it clarifies that unsolicited receipt of information and other unwitting dealings with information do not constitute 'dealing with', for the purpose of part 5.6.

These are real reforms. They reduce criminal liability, raise thresholds and lower penalties. I do not want to diminish what this bill achieves. Reducing 875 secrecy provisions to fewer than 600 is a genuine accomplishment. Removing the absurdity of journalists facing prosecution for receiving unsolicited documents is a genuine improvement. But genuine improvement is not the same as adequate protection.

First, the new general secrecy offence does not adopt a harm based approach. The replacement offence under section 122.4 applies wherever a person improperly uses or communicates Commonwealth information with an intention to obtain a 'benefit' or cause a 'detriment', terms defined in the Criminal Code dictionary as any advantage or disadvantage whatsoever. But the Human Rights Law Centre and the Centre for Public Integrity note that this departs from what the AGD's own 2023 review recommended and from what the INSLM recommended. The concern is practical; much ordinary engagement with government could theoretically involve some advantage to one party or disadvantage to another. The HRLC gives the example of civil society organisations, including the very organisations who submitted to the inquiry for this legislation, being provided with confidential draft legislation and asked to provide feedback. This is ordinary democratic participation. Whether or not it could fall within the offence's reach is, on the current drafting, at least arguable. The word 'improper' provides some guard, but it is undefined. An undefined concept of impropriety is a thin protection in a provision that carries criminal liability. That is why I support the member for Curtin's amendment, which replaces the vague concept of 'improper' communication with a concrete, harm based threshold. It limits criminal liability to communications that harm—or where a person intends to harm or is reckless as to whether the communication harms—an essential public interest, such as national security, public health or financial stability. It also explicitly protects journalism and whistleblowing that expose wrongdoing.

Second, the journalist defence remains structured against journalists. The defence in section 122.5(6), which protects public-interest journalism, has not been amended. The Alliance for Journalists' Freedom puts the position plainly: the current provision requires journalists to bear an evidential burden to establish they reasonably believed their conduct was in the public interest. That is contrary to the presumption of innocence. It is the journalist, not the Crown, who must raise, substantiate and prove the public-interest claim, often in respect of conduct the journalist may not have known was covered by a secrecy offence at the time. The HRLC, the Centre for Public Integrity and the AJF all identify this as an unimplemented reform. The INSLM's recommendation 13, which I note was agreed to in principle by the government in November 2024, was that consideration be given to recasting the defence as an exception, shifting the burden to the prosecution. That recasting does not appear in this bill. That is deeply disappointing.

Third, the Attorney-General's consent mechanism has structural limitations. I acknowledge that this is an improvement on the current arrangement, but I hold concerns about its potential for politicisation. Journalists report on politicians. Whether a potential conflict of interest is real or simply perceived, the fact that a politician holds the power to authorise or block the prosecution of a journalist for a secrecy offence is significant and should raise concern. As the HRLC noted in their submission, this mechanism is unsatisfactory as a matter of principle but, given the absence of stronger protections, desirable in practice. It is disappointing that this must be our solution in the face of reforms which don't go quite far enough. Again, I support the member for Curtin's amendment, which requires an independent review of this provision every five years in consultation with media stakeholders.

Fourth, and following on from this point, the bill has not come alongside necessary reform of the Public Interest Disclosure Act. We are still without a whistleblower protection authority; we are still without a single, consolidated whistleblower act. A parliamentary committee first called for an independent protection body for whistleblowers in 1991. It was called for again in 1994 and, again, in 2017. We are still waiting. This reform is monumental, but it does not appear in a vacuum in our legislation. It comes as David McBride is still in jail and as whistleblower advocates still await certainty. Currently, under section 122.5(4) of the Criminal Code, compliance with the Public Interest Disclosure Act can operate as a defence to a secrecy offence, but the PID Act's requirements for a qualifying 'external' or 'emergency' disclosure are highly technical. This makes it harder for whistleblowers to have certainty about whether disclosures which are well within public interest will leave them criminally liable.

I will be moving a consideration in detail amendment to address this gap. My amendment would provide that, where a whistleblower does not satisfy every technical requirement but can demonstrate that the disclosure was both reasonable and genuinely in the public interest, they may have access to additional certainty. These disclosures will still not be permitted to be given to a foreign public official. It doesn't resolve all the problems with our whistleblower framework, but it is a meaningful first step towards protections that the public expects and whistleblowers deserve.

This bill's senate committee reports on 19 June. The government's response to those recommendations will be a test of whether its commitment to further reform is genuine. I commend the bill for inquiry, and I urge the government to address these issues I have raised before its passage. Reducing this body of law from 875 provisions to fewer than 600 matters. Ending the absurdity of journalists being prosecuted for receiving documents matters. But we should not mistake progress for completion. The people this framework most affects, whether they are journalists, those who represent whistleblowers or civil society organisations, are still telling us it falls short. We should listen.

See this speech in context