Australians should prepare for online services to go down unexpectedly as threats rise, cyber agency warns

The Australian Signals Directorate (ASD) reports a significant increase in cybersecurity incidents, with responses up 20% and notifications of potential malicious activity rising 80% in the 2025-26 year. The agency explicitly links this surge to the increasing availability of AI, which allows cybercriminals and state-based actors to find vulnerabilities and launch attacks more quickly and effectively.

To counter these AI-enabled threats, the ASD advises a dual strategy. First, Australian entities must accelerate the modernization of their technology and cybersecurity foundations, specifically addressing the risks posed by 'legacy technology', outdated systems no longer supported with updates by their creators. Such systems are identified as attractive targets for hackers.

Second, the ASD advocates for integrating advanced AI tools into cybersecurity functions. These tools are deemed essential to "automate our defences to stay ahead of AI-enabled attackers," helping to identify weaknesses, detect suspicious behavior, and improve response times. As an example, the ASD itself reportedly has access to Anthropic’s Mythos model, an AI tool noted for its potential to identify unknown flaws in IT systems.

While the report outlines the critical need for both foundational modernization and the deployment of advanced AI, it does not detail the specific implementation challenges, resource allocation, or timelines required for widespread adoption across Australia's diverse organizations and critical infrastructure.