Australian Labor seeks answers on rogue AI as cyber threat cost hits $37B

The Australian Financial Review reported on September 27 that Labor is seeking answers on the risks posed by rogue AI, alongside a figure putting the cost of cyber threats at roughly $37 billion.

The story ties two threads together: the technical risk that AI systems can be misused or escape intended controls, and the cybersecurity bill that such capability amplifies. Putting a dollar figure on the threat gives policymakers a measurable anchor, moving AI risk out of abstract ethics debate and into budget and regulatory territory.

Demanding answers is itself a political act. It usually signals the issue has left academic and industry circles for an accountability track, where hearings, departmental questioning or amendments to existing rules tend to follow.

For the AI industry, the shift worth watching is in what regulators are asking about. Earlier debates centred on training data and model bias; attention is now moving toward automated attacks, the permission boundaries of autonomous agents, and whether pre-release safety evaluations are adequate.

Quantifying a risk in dollars is a common precursor to resource allocation. Once the cost is named, it becomes easier to argue for budgets, mandates and new oversight — which is why the number matters as much as the political demand around it.

What to watch next: whether the questions turn into a concrete timetable or legislative action, and whether domestic firms disclose corresponding security and compliance spending. Vendors that can present verifiable safety evaluations in advance are likely to find the conversation easier than those responding to questions after the fact.