Australia lost an estimated $87.39bn to money laundering in 2024-25, according to analysis from Napier AI, which argues that artificial intelligence must be used to strengthen anti-money laundering (AML) foundations rather than bypass them entirely.
The Financial Action Task Force’s April Ministerial Declaration confirmed that AI misuse, virtual assets and increasingly complex financial structures are now priorities for the 2026-2028 period, reflecting threats already playing out rather than looming risks.
Napier AI points to stablecoins, digital asset layering and synthetic identity fraud as typologies that traditional monitoring systems struggle to catch in real time. Trade-based money laundering also remains a persistent blind spot as digital platforms accelerate cross-border commerce without necessarily improving transparency.
Drawing on the Napier AI/AML Index, the firm highlights Australia as a case study, where human trafficking enabled through money muling networks, drug-related flows through informal systems, and smurfing at scale dominate criminal activity. Criminals are increasingly using AI to automate these schemes, turning smurfing into a high-volume, low-cost operation.
The economics are stark. While AI could potentially recover $2.65bn of the losses, money laundering losses are rising 3% year-on-year, while compliance costs are climbing 9%, well above global averages. Napier AI warns this trajectory risks tipping Australia, currently seen as an effective leader in balancing compliance costs and outcomes, into inefficient overspending.
Regulator AUSTRAC has been explicit that institutions cannot hand compliance responsibilities entirely to AI. Napier AI stresses that responsible AI adoption must begin with a risk-based assessment before deployment, determining acceptable risk levels and where automation can safely reduce noise.
A recent Federal Court of Australia judgment cautioned against using large language models to summarise complex material without human validation, reinforcing Napier AI’s view that human-in-the-loop oversight remains non-negotiable, particularly for high-risk decisions.
Napier AI’s core recommendation is that institutions reconnect AI adoption with foundational AML capabilities, since legacy platforms built on batch processing and static rules cannot be transformed by simply layering AI on top. Effectiveness, the firm argues, should be measured by detection quality and investigative accuracy rather than volume reduction alone.
For more, read the full story here.
Copyright © 2026 FinTech Global
Copyright © 2026 RegTech Analyst
