Australia created a criminal offence for creditor-defeating dispositions in 2020 carrying up to 10 years imprisonment — ASIC dated its own first charge under that offence to 2026-06-30

What is happening?

The Treasury Laws Amendment (Combating Illegal Phoenixing) Act 2020 of the Commonwealth received assent on 2020-02-17. It created a new head of liability in the corporations legislation, the creditor-defeating disposition, and it made that conduct both a civil and a criminal matter. For an individual the maximum is the greater of 4,500 penalty units, three times the benefit obtained or the detriment avoided, or 10 years imprisonment. For a body corporate it is the greater of 45,000 penalty units, three times the benefit, or 10 percent of annual turnover.

The Act was legislated against a measured harm. A 2018 study commissioned jointly by the Australian Taxation Office, the Australian Securities and Investments Commission and the Fair Work Ombudsman, and carried out by an external professional services firm, put the direct annual cost of illegal phoenixing to the economy at 2.85 billion to 5.13 billion dollars, measured on 2015-16 activity.

On 2023-06-02, three years and more after assent, a legal journal reported that only two cases had touched the provision, and that it was not clear whether ASIC had used the new powers at all.

On 2026-06-30 ASIC published a media release stating that this was the first charge brought by ASIC under the criminal creditor-defeating disposition legislation. That is six years and four months after the Act received assent. The charges laid were one under section 588GAC, which carries a maximum of 10 years imprisonment, and one under section 184 subsection 2. The matter went before a local court. No source opened here records a plea, a finding or any other outcome, and nothing in this document anticipates one.

Across that same span the civil and administrative track was running. In the 2023-24 financial year the Phoenix Taskforce, led by the Australian Taxation Office and drawing in ASIC, the Australian Competition and Consumer Commission and police agencies, recovered 99 million dollars in cash, disqualified 8 directors, received more than 3,400 reports and completed more than 2,600 audits and reviews. Its cumulative recovery since 2014 is 2.45 billion dollars. The track that produced figures every year was the civil and administrative one, and the criminal limb is the one ASIC dated to its first use in 2026.

Whose problem is this?

The agencies that commissioned the measurement of the harm are among the agencies that hold the instruments created to address it. The instrument that produced a public record every year is not the instrument this document is about.

Where does this problem end?

Three value questions sit immediately beside this document and none of them is settled in it. Whether limited liability and the winding-up system are themselves the right design is the first. Whether criminal prosecution or civil recovery does more for the people who are out of pocket is the second, and the record here contains both tracks without ranking them. Whether the penalties set in 2020 are proportionate is the third. This document measures a distance between a provision and its first use, and takes no position on which instrument ought to close that distance.

What is the state now, and what should it be?

Now

Needs a new measurementthe target state: no source opened here names a target. Not the Act, not the regulator, not the taskforce, and not the 2018 cost study. There is no published number of prosecutions the criminal provision was meant to produce, no timetable for its first use, and no figure by which the annual cost estimate was meant to fall. Nobody is failing to hit a number here, because no number was published.

How big is it?

The harm is published in dollars and not in people. The 2018 study measured direct annual cost on 2015-16 activity and gave a range of 2.85 billion to 5.13 billion dollars. Its component figures were about 300 million dollars a year in unpaid employee wages and entitlements, about 1.6 billion dollars a year in direct tax loss to government with a wider economic effect of 750 million to 1.5 billion dollars, and lost gross domestic product of up to 3.5 billion dollars, which the study expressed as 0.21 percent of annual output.

The regulator cites the same 2018 study on its own page and publishes a different set of figures — about 3.19 billion dollars in total annual harm, split into up to 655 million dollars for employees, up to 1.93 billion dollars for business and up to 610 million dollars for government. Both aggregates are attributed to the same study and they disagree at the total and at every component. Neither is adopted here as the correct one.

No source opened here converts either aggregate into a count of people. Because no headcount exists anywhere in the material, the population estimate for this document is recorded as not derivable rather than assembled from a division that no source performs.

The recovery figures do not net against the cost estimate. The taskforce total of 2.45 billion dollars since 2014 covers the whole of its activity, and the source that publishes it does not separate illegal phoenixing from the other avoidance types the taskforce works on.

Under what conditions does it arise?

1. The same conduct sits on two tracks and the tracks leave different records. The 2020 Act created civil and criminal liability in one instrument. An agency working a matter can move on either, and only one of those tracks produces a figure that appears in an annual total.

2. The civil and administrative track publishes yearly. Cash recovered, directors disqualified, reports received and audits completed are all counted and released each financial year, which makes activity on that track continuously visible from outside.

3. An unused criminal provision produces no document at all. Between assent and 2026-06-30 the regulator published no charge under the criminal limb, and a provision being held in reserve looks exactly the same from outside as a provision nobody is preparing to use. The absence became visible only when the regulator itself described a charge as the first one it had brought.

4. Judicial consideration stayed thin for years. As at 2023-06-02 one opened commentary counted two cases touching the provision and recorded that it was not clear whether the new powers had been used, which means the case law that would ordinarily settle how the provision works had barely begun to accumulate.

What has been tried?

Across the span from assent to 2026 the enforcement work that produced published figures every year was the civil and administrative one. The measurement direction was completed once in 2018 and no source opened here restates it on a later activity year.

What was found?

Why is it still unsolved?

Enforcement absent — the instrument exists and, on the account the regulator itself published, it produced its first charge six years and four months after the Act received assent.

No source opened here records a reason. Not ASIC, not the taskforce, not the Parliament, and not the 2018-19 performance audit whose conclusions were not opened. This document does not supply one. Whether the gap reflects the burden of criminal proof, the way matters are referred, the resources allocated to each track, or something else entirely is not established by any material behind this document, and the honest state of the record is that the reason is unrecorded rather than known.

What is established is what stood beside the gap. Over the same span the civil and administrative track produced 99 million dollars of cash recovery, 8 director disqualifications, more than 3,400 reports and more than 2,600 audits and reviews in a single financial year, against 2.45 billion dollars cumulative since 2014. ASIC publishes its own account of acting on illegal phoenix activity and sets criminal, civil and administrative instruments out together on that page. So the criminal limb was not withdrawn, unlisted or disclaimed. It was listed, and until 2026 it was not used by that agency.

The second part is that this state of affairs generates no signal. A track that recovers money publishes a number every year and a track that lays no charge publishes nothing, so the two are not comparable from outside and only one of them can be seen to be running. Nothing in the material opened here sets a number of prosecutions the provision was meant to produce, a date by which it was meant to be used, or a review point at which non-use would have been noticed. The gap became visible in 2026 only because the agency described its own charge as the first one, which is to say the record disclosed the six-year absence at the moment the absence ended.

What observation would mean it is solved?

Candidates — (a) charges under the provision become a recurring entry in the enforcement record rather than a single event (b) the direct annual cost of illegal phoenixing is measured again on a recent activity year and is lower than the 2015-16 estimate (c) the published figures for the civil and administrative track and the criminal track can be read side by side, so that non-use of one is visible while it is happening.

(a) alone counts inputs. A charge is not a conviction, and a second charge would show that the provision can be used rather than that it is deterring anything. This document takes no position on the outcome of any particular matter, and a run of charges that all fail would be a different result from the same observation.

(b) alone has no baseline to move against. The only cost figures in this record rest on 2015-16 activity measured once in 2018, and the two published aggregates for that single study already disagree at the total and at every component. A later study could differ from either for reasons of method rather than of harm, and no source opened here restates the estimate on a later year at all.

(c) is the weakest observation and the most useful one. It changes nothing about the conduct and only makes the state of each instrument legible while it holds. The reason it matters is in the record above: the six-year gap was disclosed by the regulator at the moment it closed, and an observation that arrives only when a condition ends cannot be used to notice the condition.

What is it connected to?

Fills with researchinsolvency and liquidation policy, employee entitlement guarantee schemes, director identification requirements, and the enforcement records of comparable corporate criminal provisions in other jurisdictions. Relation type and evidence grade were not confirmed in this round, and no source opened here draws any of these connections.

What these sources do not say

• How many people are affected. Every estimate in this record is expressed in dollars. No source opened here counts employees, creditors or companies, and none states whether such a count was attempted and failed or was never attempted.
• Which of the two cost aggregates is correct. The range of 2.85 billion to 5.13 billion dollars and the total of about 3.19 billion dollars are both attributed to the same 2018 study, and they disagree at the total, at the employee share and at the government share. No opened source acknowledges the other, states its counting basis, or says which is the later restatement. None of them restates the currency basis either.
• When the offence commenced. The sources opened here give the assent date of 2020-02-17 and none gives a separate commencement date, which in Australian practice can fall later. The span of six years and four months in this document is therefore measured from assent, and would be shorter if commencement fell later.
• Why no criminal charge was brought before 2026. No source opened here carries an on-the-record explanation from ASIC, from the taskforce or from any other body. Sources record the absence and its endpoint; none records a reason for it.
• What the 2018-19 performance audit concluded. The report title is confirmed as Addressing Illegal Phoenix Activity and its conclusions are not, because that document was not opened here. Its findings are therefore absent from this record rather than summarised from a second-hand account.
• How many civil actions were brought under the new provision. One opened commentary counts two cases as at 2023-06-02 without saying whether they were civil proceedings, criminal proceedings or something else, and no source opened here gives a count of civil actions under the provision at any date.
• What the taskforce recovery figures contain. The 2.45 billion dollars since 2014 and the 99 million dollars for one financial year cover the whole of taskforce activity. No opened source separates the portion attributable to illegal phoenixing from other avoidance types, so those figures cannot be read as a measure of this provision.
• Anything about the outcome of the 2026 matter. It is before a court and no plea, finding or sentence exists to report. No source opened here states one, and no inference about one is drawn anywhere in this document.
• Whether any suppression or non-publication order applies to that matter. No source opened here addresses the question in either direction, so the absence of such an order is also unconfirmed.

See the evidence

No primary statutory or audit document was read in full. The two documents opened directly from the body this record is about are both published by ASIC — the media release dating the first charge, and the page setting out the instruments it uses on illegal phoenix activity. The statute itself and the 2018-19 performance audit were not opened, so the penalty structure, the assent date and the section numbers all rest on professional and trade sources that cite them. Where those sources overlap they agree: the assent date of 2020-02-17 appears in two, the maximum of 10 years imprisonment for an individual appears in three including the summary of the 2026 charge, and the description of the 2026 charge as the first under the criminal provision appears in two, one of which is the regulator itself. Where they disagree the disagreement is left visible rather than resolved — the 2018 cost study is reported as a range of 2.85 billion to 5.13 billion dollars by one opened source and as a total of about 3.19 billion dollars by the regulator, with the employee and government components differing between them, and no opened source acknowledges the other or states its counting basis. This is a Path A output, so observation_refs is empty and provenance_mode: press-derived.

People affected

Estimated range Not derivable

The reason and what is missing are listed under “What is missing” below