Australia’s online betting industry is getting smarter. That may not be entirely good news.
ACMA has released new research into how artificial intelligence is being used across online gambling services accessed by Australians. The short version: licensed wagering providers are already using AI in ways that shape odds, promotions, customer experience and risk detection. The longer version is where things get uncomfortable.
The regulator’s report breaks the current use of AI into four main buckets: predictive analytics and odds setting, personalised promotions and services, content and product design, and detection of harmful or fraudulent behaviour. That covers a lot of ground. It is not just a chatbot in the help centre or a marketing team playing with image prompts. It is the machinery behind how betting platforms price markets, nudge users and manage risk.
For operators, the attraction is obvious. AI can process live data quickly, build more detailed odds, tailor the app to each customer and flag suspicious behaviour faster than older systems. It can also help spot patterns linked to gambling harm, which gives the technology a genuine safer gambling use case.
That is the clean version of the sales pitch.
When Smarter Betting Tools Create Harder Questions
The harder question is what happens when the same technology that can identify risk is also used to keep people betting for longer.
ACMA’s report notes that AI-supported personalisation can shape what users see, which promotions they receive and how long they remain on a platform. Gambling companies often call that engagement. Harm researchers tend to use less cheerful words when the person being engaged is already in trouble.
That tension is not new in gambling. It is just getting faster.
A bookmaker has always wanted to show customers the markets most likely to interest them. AI makes that easier, sharper and more individual. One customer might see racing offers. Another might be pushed towards same-game multis. Someone else might get a homepage filled with events that match their recent betting habits. None of that has to look sinister on the screen. It can look convenient.
Convenience is not neutral when money is involved.
The report also points to AI’s role in odds setting. Predictive systems can use live information such as injuries, weather and betting activity to generate or adjust odds quickly. They can also create markets on more specific outcomes, such as player stats or smaller in-game events. More markets mean more choice. They also mean more ways to bet.
That matters in Australia because online wagering is already a major part of the gambling market. ACMA’s report cites estimates that Australians spent $12.45 billion on online gambling in 2024, with 70% of that spending through wagering. It also notes online casino gambling remains illegal in Australia under the Interactive Gambling Act.
So this is not a tiny tech issue sitting off to the side. It is happening inside one of the biggest legal online gambling channels available to Australian customers.
There are benefits. AI can help identify fraud, money laundering signals, account misuse and suspicious betting patterns. It can scan transactions and behaviour at a speed human teams cannot match. For operators dealing with AML obligations, multi-account abuse and integrity concerns, that is useful.
It can also help with safer gambling if used properly. Models can pick up changes in deposit patterns, session frequency, betting intensity and other signs that a customer may be moving into harmful play. A human might miss the pattern. A well-trained system should not.
The key words are “if used properly”.
AI does not come with a built-in conscience. It follows the goals set for it. If the target is safer gambling, it may help spot risk earlier. If the target is more activity, more deposits and more turnover, it may become a very efficient nudge machine.
That is why governance matters. ACMA said AI adoption is delivering efficiencies across gambling, media and telecommunications, but is also prompting calls for stronger safeguards, transparency and oversight. In gambling, that cannot be treated as a future problem. The tools are already here.
Why AI Oversight Cannot Wait for the Next Scandal
The most awkward part for operators is explainability. If an AI model decides a customer should see a certain offer, be shown a certain market or be flagged for intervention, someone inside the company needs to understand why. “The model said so” is not a responsible gambling policy. It is barely an excuse in a group chat.
Australian wagering companies are already under pressure over advertising, self-exclusion, offshore competition and financial crime controls. AI now sits across all of those areas. It can personalise ads, support BetStop-related account checks, detect suspicious payments and build new product features. Used well, it can strengthen the system. Used badly, it can make every existing problem harder to see until it is larger.
Regulators will also have to keep up. Traditional gambling rules were not written for platforms that can change the customer experience in real time, customer by customer. A blanket rule about marketing or inducements may be harder to enforce when the offer, timing and placement are all being decided by automated systems.
That does not mean AI should be banned from betting. A ban would be both unrealistic and probably unhelpful. The better question is what operators should be allowed to optimise for.
If AI is used to detect harm, stop fraud and improve compliance, most people will see the value. If it is used to identify vulnerable customers and serve them sharper promotions, the public reaction will be very different. Nobody wants a betting app that knows someone is spiralling and decides the answer is a better bonus.
ACMA’s report is not an enforcement action. It does not accuse operators of breaking the law. It is a warning map. It shows where the technology is moving and where regulators may need to look next.
For Australian players, the change may be almost invisible. The app feels smoother. The offers look more relevant. The markets appear faster. The support bot answers quickly. That is how AI usually arrives: not with a label, but with the small feeling that everything is suddenly tailored.
In gambling, tailored can be helpful. It can also be dangerous.
The industry will say AI can make betting safer and more efficient. That can be true. The test is whether safer gambling gets the same energy as sharper marketing and better margins.
Because if AI is going to help shape the future of Australian wagering, the first question should not be how clever it is.
It should be whose side it is on.