Earlier this week, our CEO and Co-founder
@theonejvo spoke with
@MizenRonald, Senior Political Correspondent at the The Australian Financial Review, for a piece on Assistant Technology Minister
@Charlton_AB's push to keep Australian AI procurement spend onshore.
The framing of the piece is a useful one for the future.
Every AI procurement decision Australian businesses make today have small individual consequences and enormous collective ones.
$100 billion in wealth retained by developing and exporting Australian AI versus the "Uberisation" of the digital economy with profits flowing offshore, is not an abstraction, it is the actual choice Australian boards are making in real time, often without recognising it as a choice at all.
That argument applies most sharply to AI cyber offense and defence capability, which is where Aether AI sits.
@tryaether_ai is Australia's, and the world's, first full-spectrum frontier attack and defence AI, securing cloud, mobile, wireless, web, identity, API and adjacent surfaces under a single architecture.
Cyber is no longer a tool category that can be outsourced as a procurement convenience. It is becoming a national security & capability question.
@AnthropicAI's Mythos changed the threat picture, and project Glasswing's invite-only access list, with no Australian bank publicly listed on it, illustrated that foreign-controlled frontier AI is now a strategic asset held by its host country, available to other countries on terms that can be influenced by the host country's politics.
None of this is opposition to Anthropic.
In fact, we've been accepted into their Cyber Verification Program, use their frontier models inside our architecture (alongside our own cyber model), and we want to keep working alongside them on what comes next. So, the sovereignty argument is structural, not adversarial.
Without it, a meaningful part of the country's cyber posture is outsourced.
AUKUS Pillar II was designed to make Australia a co-developer of advanced capability, not a customer.
On AI cyber specifically, that role only works if Australia has its own offensive and defensive AI capability to bring to the trilateral, therefore, if the country does not build it now, the next decade of Five Eyes cyber cooperation will be entered as a buyer. That is not the role AUKUS was designed for.
Last week's coverage on Mythos, the
@ASDGovAu's updated frontier-AI guidance, and the
@AUSTRAC and
@APRAinfo commentary, describe the same thesis from four different angles.
1. The AI cyber threat picture has shifted
2. Australian organisations are not adequately positioned
3. Foreign capability cannot be the default answer
4. The window for Australia to build its own AI capabilities is short and closing.