News analysis · Safer gambling
Australia’s gambling survey puts online safeguards to the test
ANU research finds persistent indicators of risk alongside stable gambling participation. New controls will need to deliver measurable protection for players and households.
In this article

Related coverage: our reporting on Australia’s wagering advertising framework.
Stable gambling participation is giving Australia little reassurance on harm. New research from the Australian National University finds persistent indicators of risk while online play remains widespread, putting the effectiveness of consumer protection back in focus.
Publicised by ANU on 13 November, the report by Aino Suomi, Markus Hahn and Nicholas Biddle draws on seven survey waves since 2019. The latest survey was conducted in March 2026 among 3,482 adults, with results weighted to Australian population benchmarks.
For operators and policymakers, the important question is what a stable participation figure can conceal. The number of people taking part may change little while the products they use, their frequency of play and the consequences for their households change considerably.
What the headline figures measure
| Measure | Share of Australian adults |
|---|---|
| Gambled in the past year | Just over 60% |
| PGSI score of one or more | Around 17% |
| Gambled online at least half of the time | 35% |
| Negatively affected by someone else’s gambling in the preceding year | 5.1% |
The 17% measure uses a score of at least one on the Problem Gambling Severity Index. The standard PGSI categories distinguish scores of one to two, indicating low risk, three to seven, indicating moderate risk, and eight or more, the problem gambling category.
The headline therefore covers a broad range of reported experiences. Treating everyone within it as having the same severity would obscure the finding. It is more useful to ask where difficulties are emerging and whether protection reaches people before those difficulties become more serious.
Online access changes the practical setting
ANU describes participation as having stabilised since 2023, following sharper declines during and immediately after the pandemic. Lottery tickets remain the most common activity. Online gambling is associated with greater frequency and higher rates of at-risk gambling in the research.
That association needs careful interpretation. The survey cannot by itself establish that moving online caused the higher risk. People already gambling more frequently may be particularly likely to use online services. Both possibilities matter when deciding which interventions to test.
The shift also reaches products outside Australia’s permitted online offer. ANU says the proportion of casino table-game players reporting online use rose from 5% in 2024 to 22% in 2026. Among poker-machine players, the corresponding proportion rose from 7% to 13%.
ACMA explains that the Interactive Gambling Act prohibits providers from supplying online casino services to people in Australia.
Our concern is the gap between a rule and a customer’s actual experience. Protection depends on whether prohibited services remain reachable and whether safeguards on permitted services work during ordinary use. A legal distinction has limited protective value if the consumer cannot recognise it or act on it.
The consequences reach beyond the account holder
ANU’s affected-others measure asks about negative effects from another person’s gambling over the previous year. It overlaps with other groups in the survey; the percentages cannot be added to produce a total number of people experiencing harm.
This widens the assessment beyond an individual account. Household money, relationships and the time available for family life can be affected even when another family member never opens a gambling app. A participation statistic alone cannot describe those consequences.
For the industry, that means the quality of a customer relationship should include how the operator responds to signs of difficulty. A business can report stronger engagement while a household experiences that same engagement as a growing burden. The survey does not allocate that burden to particular brands.
The policy history gives January a harder test
The debate predates this survey. In June 2023, the parliamentary inquiry chaired by Peta Murphy produced 31 recommendations on online gambling harm. They included a comprehensive advertising ban phased in over three years, stronger consumer protections and better research and data.
The forthcoming framework includes restrictions on online wagering advertising from 1 January 2027. ACMA’s November guidance sets out a triple lock exception involving protection of restricted users, registered account use and an advertising opt-out. Additional national register obligations apply when that register commences.
We examined the implementation in our earlier report on ACMA’s triple lock. The ANU findings add an outcomes question: what evidence would show that these controls have made everyday gambling safer?
The March survey provides an earlier reference point. It cannot evaluate measures that begin the following January. Future comparisons will need consistent questions and methods, with enough time for changes in exposure and behaviour to emerge. A fall in advertising impressions would answer only part of that assessment.
Our view: judge protection by what it changes
We would judge the reforms by whether people refusing gambling advertising actually receive less of it, whether operators respond promptly to signs of difficulty and whether households report fewer negative consequences. Compliance records can help establish what happened, but they need to connect with those experiences.
The commercial test is equally practical. An operator should be able to show that its protection tools remain easy to use when they interrupt a profitable customer journey. Limits, exclusions and advertising choices lose value if the customer has to navigate avoidable friction to make them work.
Australia’s next survey results will deserve attention even if participation remains broadly unchanged. The more revealing movement would be in the severity of reported difficulties and the consequences for other people. That is where a stronger framework needs to demonstrate its value.