Damian Hickey initiated the Casuarina Foundation after watching the world's leading AI model companies push AI safety down the priority list in favour of competitive speed to market. Despite an accelerating global conversation about AI risk, no viable technical solution had emerged that could enforce governance at a level the AI system itself could not override.
The EU AI Act created binding legal obligations for AI governance, but no technology existed to meet them. Governments were writing regulations with no enforcement architecture. Industry was offering voluntary commitments with no structural accountability. And the fundamental technical challenge, converting probabilistic AI outputs into deterministic, auditable, and compliant results suitable for high-risk and regulated use cases, remained entirely unaddressed.
Mr. Hickey's response was not a policy paper. It was a patent. Filed with the United States Patent and Trademark Office in January 2026, the Deterministic Governance Architecture establishes a hardware-enforced governance pathway that the AI system cannot bypass, suppress, or decline to invoke. Governance is triggered not by software decision, but by a dedicated hardware signal responding to the semantic state of the AI model itself.
He then established the Casuarina Foundation in Australia, deliberately structured from inception as a coalition steward rather than a single-portfolio mission entity, and constituted as a Company Limited by Guarantee. The Triodian portfolio enters as Coalition Member 1 under the Coalition Founder Undertaking Deed, the Coalition Head Licence, and the Master Licence Agreement, the same instruments every subsequent coalition member will execute. The choice was structural, not symbolic. A single-portfolio steward, however well-intentioned, can be perceived as serving one inventor's interests. A coalition steward administering identical terms to every contributor is structurally different.
The choice of jurisdiction was equally deliberate. An Australian mission entity is domiciled outside the political volatility of any single AI superpower, supervised by an independent Commonwealth statutory regulator, and reads as globally credible to non-US audiences. A separately-incorporated US 501(c)(3) sister affiliate operates alongside the Foundation to receive US-tax-deductible donations and run public-benefit grant programmes globally.
The Foundation is not a gesture toward responsible AI. It is the structural instrument through which AI safety and AI governance technology, contributed by Triodian and by every subsequent coalition member, whether hardware-enforced, software-based, or hybrid, is made permanently available as a public good.
