Abigail Bradshaw, director-general of the Australian Signals Directorate (ASD), warned on Monday, 15 September 2026 that Australia has no formalised early warning system for threats posed by artificial intelligence — a gap she described as increasingly urgent as AI-driven attacks begin targeting outdated government and business infrastructure. The warning came at a summit held in Canberra, underscoring the growing anxiety among Western intelligence agencies about AI's dual-use potential.

The Warning from Australia's Cyber Intelligence Chief

Bradshaw drew a direct contrast between Australia's mature cyber threat architecture and its nascent AI risk framework. 'What we are missing at the moment, which we have in a cyber context but we don't yet have it formalized in an AI context, is an early warning system,' she said at the summit.

She urged organisations to adopt AI as a defensive tool rather than treating it solely as a source of risk. Critically, she called for intelligence and security agencies to partner directly with AI companies to identify and flag emerging threats before they are weaponised — a step that would require formal mechanisms currently absent in Australia's policy landscape.

Global Context: Industry Leaders Call for a Slowdown

The Canberra summit followed a significant moment in the international AI safety debate. Just days earlier, on Saturday, Anthropic chief executive Dario Amodei publicly called for the industry to slow its development pace and for independent, third-party evaluators to verify that AI models comply with safety standards. Both OpenAI CEO Sam Altman and xAI founder Elon Musk said they agreed with Amodei's position — a rare moment of apparent consensus among competitors.

This chorus of concern from industry insiders has intensified pressure on governments worldwide to move beyond voluntary commitments and establish binding oversight frameworks. Australia's intelligence community appears to be pushing in the same direction, albeit from a national security rather than a commercial angle.

Australia's Government: Don't Leave It to Companies or Superpowers

Australian Assistant Minister for Science, Technology and the Digital Economy, Andrew Charlton, responded pointedly to the industry's calls for self-regulation. Speaking at the same summit on Monday, Charlton distanced the Australian government from the idea that either technology companies or major powers should dictate global AI guardrails.

'I don't think we should care very much about what individual chief executives say or what individual executives claim they want to happen,' he said. 'I am not in favour of leaving it to the companies and I'm not in favour of leaving it to the superpowers.'

The remarks signal that Canberra is seeking a more multilateral, sovereign approach to AI governance — one that does not cede standard-setting authority to Washington, Beijing, or Silicon Valley boardrooms.

Why the Early Warning Gap Matters

The absence of a formalised AI early warning system leaves governments and critical infrastructure operators without the equivalent of the threat-intelligence feeds that now underpin cybersecurity responses. In the cyber domain, agencies like the ASD routinely share indicators of compromise across sectors; no comparable architecture exists for AI-specific attack vectors, such as model manipulation, adversarial inputs, or AI-assisted social engineering at scale.

Notably, this is not a uniquely Australian problem — similar gaps have been flagged by intelligence communities in the United Kingdom, United States, and across the European Union. However, Australia's public articulation of the deficit at a senior intelligence level marks a step toward making the gap an actionable policy priority rather than an abstract concern.

What Comes Next

Bradshaw's call for a formal AI early warning mechanism is likely to feed into Australia's evolving AI regulatory agenda. The summit's conclusions are expected to inform ongoing inter-agency discussions on whether such a system should be housed within the ASD or a new dedicated body. How quickly Canberra translates the intelligence community's warnings into concrete governance architecture will be the real measure of urgency.