Australian Intelligence Warns AI-Powered Cyberattacks Could Incapacitate Nation

The Australian Signals Directorate (ASD) warned that frontier artificial intelligence models have fundamentally redefined the cyber threat landscape, creating risks that could potentially incapacitate Australia. In its annual threat report, the agency revealed a 20 percent increase in cybersecurity incidents during the 2025-26 financial year, responding to over 1,500 events and issuing more than 4,400 notifications of malicious activity—an 80 percent surge.

ASD Director-General Abigail Bradshaw and Deputy Prime Minister Richard Marles warned that state-sponsored actors from China, Russia, and North Korea, including groups such as Volt Typhoon and Lazarus Group, are using AI to identify software vulnerabilities and preposition themselves within critical infrastructure. Officials highlighted that these capabilities could be used to target superannuation savings totaling over $4.5 trillion. To counter these threats, the government is urging organizations to prioritize system security over availability, meaning essential services may be taken offline unexpectedly to be secured.

While the ASD reported offensive successes, including the dismantling of the Amadey malware network which protected 180,000 computers, the agency emphasized the need to modernize legacy technology. This urgency is echoed by Moody's Ratings, which noted that while AI can weaponize vulnerabilities in days, the median rated company takes 87 days to remediate half of known exploits. In response, the ASD has established early warning protocols with Five Eyes intelligence agencies and AI developers like OpenAI.