Overview
- The Australian Signals Directorate’s annual threat report, which was released Monday, recorded more than 4,400 suspicious activity notices and over 1,500 incidents in 2025–26, an outsized jump from the prior year.
- The report names China, Russia and North Korea as principal state threats and says those actors are using AI to preposition inside networks in ways that could enable strategic disruption of services.
- Frontier AI models have been used to automate vulnerability hunting and exploit chaining, with recent discoveries of flaws in PHP forcing organisations including the Australian Bureau of Statistics to take sites offline for emergency patching.
- ASD leaders and financial regulators are urging urgent measures such as removing exposed credentials from code histories, enforcing multi‑factor authentication, prioritising patching over short‑term availability, isolating test agents from production, and adopting AI tools for automated detection.
- Officials warn the economic stakes are high because attackers could target banking systems or the AU$4.5 trillion superannuation sector, raising the risk of account loss, higher business disruption costs, and expanded regulatory probes of AI services.