How often are migration agents actually sanctioned?

The 61 checks out, and here is the working

The ministerial claim is that 61 registered migration agents have been sanctioned since 2021-22, of which 14 came in the current financial year.

Each annual figure is published separately in the Department of Home Affairs annual reports and OMARA's own activity reports. Laid end to end:

2021-22: 13 agents sanctioned. 2022-23: 4. 2023-24: 10. 2024-25: 20. 2025-26 to date: 14.

That totals exactly 61. The government's headline number is independently verifiable from the government's own documents, which is more than can usually be said of an enforcement statistic.

One qualification matters. The 14 for 2025-26 is a part-year figure, covering 1 July 2025 to 17 March 2026, the date of the ministerial release. It is not a completed financial year and should not be compared like-for-like against the full years before it. On the run rate implied, the completed year would land higher.

The five-fold rise is real, but the low year is partly an artefact

Going from 4 sanctions in 2022-23 to 20 in 2024-25 looks like a fivefold enforcement surge, and that is how it is usually reported. Half of that is real and half is a change in how OMARA handles complaints.

OMARA explains the 2022-23 trough itself: the reduction "is attributable to the OMARA's focus on the design and implementation of a revised triage and assessment model during this period. Under the revised model, less serious misconduct matters are dealt with through an early resolution model." In other words, part of the drop is misconduct being redirected out of the formal sanctions process rather than misconduct not occurring.

The rise at the other end is better evidenced. The department attributes the 2024-25 increase to "an increase in OMARA's investigative resources, following the 2022 decision to expand the OMARA and uplift its regulatory capabilities" — a resourcing decision with a visible output two years later.

The honest summary is that enforcement capacity has genuinely increased, and that a chart running 4 to 20 without the methodology note overstates the change.

What the sanctions actually were

OMARA has four sanction types: cautioning a registered agent, suspending registration, cancelling registration, and barring a former agent from re-registering.

Across the four fully published years — 2021-22 to 2024-25, covering 47 sanctions — cancellation was the most common outcome at 18 cases (38%), followed by suspension at 12 (26%), barring at 9 (19%) and caution at 8 (17%).

That distribution is worth pausing on. The most severe available sanction against a practising agent, cancellation, is also the most frequently used. This is not a regime that mostly issues warnings.

The type breakdown for the 14 sanctions in 2025-26 has not been published.

One technical point for anyone computing a rate: barring applies only to former agents — people who have already left the register — so barrings are not drawn from the population of currently registered agents and should not be included in a "percentage of agents sanctioned" calculation.

How big is that against the size of the profession?

A sanction count without a denominator says nothing. At 30 June 2025 there were 5,337 registered migration agents, up 5% from 5,081 a year earlier.

Against that base, the 20 sanctions in 2024-25 represent about 0.37% of the profession. Excluding the three barrings, which applied to people no longer registered, 17 sanctions fell on currently registered agents — roughly 0.32%, or about one in 314.

The register itself has been growing. It declined gently from 4,971 in June 2021 to 4,839 at the end of 2022, then grew steadily to 5,337 by June 2025 — a 10.3% increase in two years. So sanctions roughly doubled over a period when the profession grew about 5% a year. The intensification is not simply a larger denominator producing more cases.

Complaints have risen too, though less sharply: 252 in 2021-22, 299 in 2022-23, 361 in 2023-24 and 363 in 2024-25.

A counting trap worth knowing about

OMARA publishes two different numbers that both look like a sanction count, and they differ by a factor of three.

In 2024-25, 60 complaints "resulted in a sanction decision". In the same year, 20 registered and former agents were sanctioned. Both figures are correct. A single agent can be the subject of many complaints, so one sanction decision can close out several.

The minister's 61 is on the people basis. Reporting the complaints figure as "60 agents sanctioned" would inflate the count roughly threefold, and the two numbers sit close together in the same tables.

For completeness, the 267 complaints finalised in 2024-25 resolved as: 135 dismissed, 60 resulting in a sanction decision, 36 addressed directly with the agent, 24 suspected breach notices issued, and 12 where no breach was found.

The record exists, but it is not published as data

OMARA runs a public search of disciplinary decisions. It can be searched by name, by MARN, by date and by sanction type, and each decision has its own summary page. For a member of the public checking one agent, it works.

As a dataset, it does not exist. The search results are rendered entirely in the browser, so nothing appears in the page source. There is no export, no API and no bulk download. Individual decisions are addressed by an internal database identifier rather than by MARN, so a decision page cannot be reached from an agent's registration number. The public register of agents has the same architecture and the same limitation.

The practical consequence is that the two datasets a researcher would want to join — who is registered, and who has been sanctioned — are both published, both searchable one record at a time, and not joinable at any scale without scraping the regulator's portal.

That is why this study reports aggregate patterns from official publications rather than an analysis of individual decisions. We deliberately have not scraped and cross-referenced individual sanction records: the aggregate picture is publishable from primary sources, and the individual picture would require building a database the regulator has chosen not to publish.

The statistical series stopped two years ago

This is the finding we did not expect.

OMARA's Migration Agent Activity Report is the publication designed to document exactly this — sanctions, complaints, register size, and anonymised case summaries of every disciplinary outcome. The page describing it still states: "Every 6 months, the OMARA publishes an activity report on registered migration agents."

The most recent one covers 1 January to 30 June 2024. As at August 2026, roughly two years of six-monthly reporting has not appeared.

The data has not vanished — the 2024-25 sanction figures survive in an annual report appendix, and the 2025-26 figures exist in a ministerial media release. But the dedicated series that would let anyone track enforcement systematically has lapsed, during precisely the period the government is publicising an enforcement surge. OMARA's standalone annual reports, a separate publication, stopped after 2013-14.

So the enforcement record is being announced faster than it is being documented.

The 2026 rules: a remake, not an overhaul

New regulations for registered migration agents commenced on 1 April 2026, replacing rules dating from 1998. This has been widely described as the first comprehensive overhaul of the profession's regulation in nearly three decades. That framing overstates it.

The reason for the change is sunsetting. Under the Legislation Act, instruments automatically repeal after a set period, and four instruments governing agents and continuing professional development were due to lapse on 1 April 2026. OMARA's own description is that the replacements "have been amended to include contemporary language and updated references".

What did substantively change is worth knowing. Agents must now complete two specific mandatory professional development activities each registration period — one on ethical standards, one on the Code of Conduct — though the total requirement remains 10 points, unchanged. A cap was introduced of six points of online activity in any 24-hour period; previously all ten could be completed in a single day. Entry requirements also shifted: one university's qualification was added to the approved list and another removed, the accepted English tests were widened to six, and test results are now valid for three years.

Two things did not change that were widely expected to. Registration fees were flagged to the profession during consultation and then dropped, with the department deciding they would be "considered by the Australian Government as part of a future budget process". And infringement notices were not introduced — that regime already existed for unregistered practice, and the only change is that a notice must now be issued within 12 months of the alleged offence.

We have seen claims that the new regulations introduce civil penalties or codify minimum professional indemnity insurance. We could not find a primary source for either and are not reproducing them.

Registered agents and 'scam agents' are two different enforcement systems

Coverage of migration advice regularly merges two populations that are legally distinct, and the distinction is jurisdictional rather than a matter of practice.

OMARA regulates registered migration agents. Its powers are caution, suspension, cancellation and barring. Its population was 5,337 people at June 2025, and the 61 sanctions apply to them.

Unregistered people who charge for immigration assistance commit an offence under the Migration Act, and OMARA has no jurisdiction over them at all. As the department puts it, OMARA "has no jurisdiction to investigate non-registered immigration assistance; this is within the purview of the Australian Border Force". Enforcement there means infringement notices, prosecution, visa cancellation, detention and removal.

So the Border Force operation reported in September 2025 — seven people located, removed or detained, who between them had facilitated more than 470 Protection visa applications and charged clients up to $1,422,000 — involved people who were not registered agents and could not have appeared in OMARA's sanction figures. A separate operation in January 2026 in Far North Queensland detained four people including a suspected scam agent alleged to have charged Pacific Australia Labour Mobility workers for invalid Protection visa applications, in some cases without their knowledge.

Combining the two counts would be wrong in both directions. It would also miss that registered agents are not the whole of the lawful market: Australian legal practitioners may also charge for immigration assistance and are not regulated by OMARA, so 5,337 is not the total number of people who may lawfully advise.

Methodology, limits, and how to cite this

Sanction counts by year come from the Department of Home Affairs annual reports for 2021-22, 2022-23 and 2024-25, from OMARA's Migration Agent Activity Report for January to June 2024, and — for 2025-26 — from the Assistant Minister's media release of 17 March 2026. Register sizes come from the activity reports and the 2024-25 annual report. Regulatory changes come from OMARA's legislative updates pages and the Federal Register of Legislation.

Four limits should travel with this data. The 2025-26 figure is a part year to 17 March 2026. The 2022-23 trough reflects a change in triage methodology as well as a change in conduct. The type breakdown for 2025-26 is unpublished. And barrings apply to former agents, so they should be excluded from any rate expressed as a share of the current register.

One derived figure is offered with a caution: a total of 11 sanctions in 2020-21 can be reconstructed by summing the comparative figures given in the 2021-22 annual report, but no source states it as a total, so it sits outside the verified series and outside our dataset.

We have not analysed individual disciplinary decisions or named individual agents. The aggregate record is fully supported by primary sources; the individual record would require scraping a regulator's portal that deliberately publishes no export.

Sanction counts by year, sanction types, and the register size series are all available as CSVs. Journalists and researchers are welcome to republish the tables with a linked attribution.

How many migration agents have been sanctioned in Australia?

61 registered migration agents have been sanctioned since 2021-22: 13 in 2021-22, 4 in 2022-23, 10 in 2023-24, 20 in 2024-25, and 14 in 2025-26 up to 17 March 2026. The figure reconciles exactly against the Department of Home Affairs' own annual reports.

What sanctions can OMARA impose on a migration agent?

Four: cautioning a registered agent, suspending registration, cancelling registration, and barring a former agent from re-registering. Across 2021-22 to 2024-25, cancellation was the most common outcome at 38% of sanctions, ahead of suspension at 26%.

What proportion of migration agents get sanctioned?

In 2024-25, 20 agents were sanctioned against a register of 5,337 — about 0.37%. Excluding barrings, which apply to people who have already left the register, 17 sanctions fell on currently registered agents, roughly one in 314.

How do I check whether a migration agent has been sanctioned?

OMARA publishes a searchable list of disciplinary decisions on its portal, which can be searched by name, MARN, date or sanction type. There is no bulk export or API, so checking must be done one agent at a time.

Is a scam migration agent the same as a sanctioned registered agent?

No, and they fall under different authorities. OMARA regulates registered agents and can caution, suspend, cancel or bar them. Unregistered people charging for immigration assistance commit an offence under the Migration Act, and OMARA has no jurisdiction over them — that is the Australian Border Force's responsibility, with outcomes including prosecution, detention and removal.

What changed for migration agents on 1 April 2026?

The 1998 regulations were remade because they were due to sunset. Substantively: two mandatory professional development topics were introduced (ethical standards and the Code of Conduct), online study was capped at six points in any 24-hour period, the approved qualification and English test lists changed, and test results now stay valid for three years. Registration fees were flagged for change and then left unaltered.