AI-powered fraud threats surge for Australian firms

Australian organisations across financial services, telecommunications and eCommerce are reporting higher fraud losses and growing concern that existing defences are falling behind, according to research commissioned by Experian and conducted by Forrester Consulting.

The Australian findings are based on a survey of 109 senior fraud decision-makers. It found that 65% of organisations recorded an increase in fraud losses year on year. Looking ahead, 68% expect fraud attacks to rise again in 2026 compared with the previous year.

The results suggest a widening gap between fraud attempts and the tools used to detect and block them. In Australia, 73% said their current fraud technology cannot keep pace with threats. Another measure in the study found 68% of business leaders believe their security tools are no longer adequate.

Generative AI

The research highlights a shift in the threat landscape, with generative AI emerging as a key concern for Australian respondents. Some 61% said it is the single biggest fraud threat they have ever seen, and 64% said their organisation has seen a noticeable increase in fraud attempts powered by generative AI.

Many also believe the trend is still in its early stages. In the Australian results, 58% agreed organisations have only seen "the tip of the iceberg" when it comes to AI-powered fraud.

The global study surveyed 979 senior fraud decision-makers across nine countries: Australia, Denmark, Germany, India, Italy, New Zealand, Norway, South Africa and Spain. Participants worked in financial services, telecoms and eCommerce.

Patterns varied by sector in the global findings. The research points to rising levels of social engineering and identity theft in financial services and telecoms. In eCommerce, it highlights an increase in friendly fraud and refund abuse.

Investment priorities

Many organisations are exploring new data sources and detection methods. In Australia, 67% plan to invest in device and behavioural data.

Across the international sample, technology spending is rising faster than investment in analysts. Globally, 71% said they are investing more in fraud technology than in human analysts, pointing to limits in manual reviews and rules-based systems as volumes rise and techniques change quickly.

The study also links slow progress to procurement and implementation challenges. It highlights "build-vs-buy" decisions that can delay upgrades while organisations weigh internal development against external suppliers-an added risk when attacks and tactics evolve rapidly.

Machine learning

Reviews of fraud solutions are a top agenda item for the year ahead, based on the global data. Around seven in ten respondents ranked them as their highest priority. Other focus areas include migrating systems to the cloud and investing in new tools.

The research also examines the use of machine learning in fraud detection. Among machine-learning users in the global sample, 68% reported measurable improvements in detection accuracy after implementation.

Respondents also cited real-time detection as a key advantage of machine learning (67%), along with the ability to retrain models on new data as tactics change. Another global measure found 82% agree machine learning helps detect fraud that rules-only systems would miss.

Sharing intelligence

The study points to growing interest in sharing information between organisations. Globally, nearly 73% agreed that sharing fraud intelligence is important for staying ahead of threats, and three-quarters said building trust within fraud consortiums is essential for success.

Experian is developing centrally managed hubs with API connections to share data across networks while meeting regulatory requirements.

Mathew Demetriou, Managing Director, Software Solutions A/NZ at Experian, said Australian organisations are reassessing their approach as losses increase and the threat environment changes.

"Australian organisations are adapting to a rapidly evolving fraud landscape," Demetriou said. "Fraud losses are rising, attacks are becoming more sophisticated, and many businesses know their existing technology can't keep pace. The rise of generative AI is accelerating both the scale and sophistication of attacks, making it critical for organisations to reassess their fraud prevention strategies. Device intelligence, behavioural analytics, and machine learning are becoming essential to effective defence. The organisations that act now to modernise and collaborate will be the ones best positioned to protect trust and growth in the years ahead."

The report points to further investment in fraud technology and data-led detection over the next planning cycle, as Australian respondents anticipate another increase in attacks in 2026.