Australia calls AI an 'unprecedented' cyber threat: number of warnings up 80%

AI has changed the cyber threat landscape, Australia says

The Australian Signals Directorate on Tuesday published its annual cyber threat report, in which it called artificial intelligence the 'defining force' in this field over the past year. According to the agency, AI allows malicious actors to find vulnerabilities faster and exploit them with 'unprecedented scale, speed and sophistication', while criminals and state-backed hackers are conducting attacks 'on a greater scale than ever before' (this is reported by AFP).

A sharp rise in the number of incidents

The report recorded a significant increase in cyber incidents: over the past year, the directorate issued more than 4,400 warnings about potentially malicious activity — 80% more than in the previous 12 months. The authors noted that state-backed cyber actors continue to target Australian organisations in order to gain access to sensitive or non-public information, and that this activity extends beyond government networks, affecting critical infrastructure facilities, private businesses and private individuals. The report also said that hackers linked to China, Russia and North Korea have recently attacked Australia or its allies.

New threats and a call to protect systems

The report singled out a new phenomenon: 'in recent weeks, threats have emerged from the unintended actions of AI agents and the testing of models'. This echoes growing global alarm following a series of hacking incidents involving models from leading developers such as OpenAI and Anthropic. At the same time, the directorate itself acknowledges the technology's dual nature: it used some AI tools in compiling the report, noting that rapidly evolving models can bring enormous benefits when applied correctly. Defence Minister Richard Marles told journalists that 'AI is already here, and it is real. People need to get on board', adding that 'bad actors are using AI to find vulnerabilities in systems'. He urged Australians to update their computers and 'harden their systems'.

A specific example: a vulnerability found with the help of AI

The practical backdrop for the warnings was a recent situation in which governments and companies rushed to fix critical vulnerabilities in a programming language that powers a significant portion of the world's websites. As AFP notes, the vulnerability was discovered using a powerful OpenAI model — an example of how the very tool whose risks the report warns about also works in favour of security.