The Quad is Australia, India, Japan and the United States. Its leaders met in Manila on 17 November, and the Prime Minister’s site publishes their joint press release. One paragraph of it is about AI, though it never uses the word:

We affirm the importance of the Super Intelligence* Accelerator, announced by our Foreign Ministers in Atlanta. The Accelerator is intended to coordinate and fund efforts to support trusted Super Intelligence (SI)* access, deployment, and technical assistance across the Indo-Pacific.

Quad Leaders’ Summit Joint Press Release, 17 November 2026, as published on pm.gov.au

The asterisk leads to the last line of the release: In Australia, India, and Japan, ‘Super Intelligence’ refers to ‘artificial intelligence.’ The White House copy of the release carries the same footnote. Outside that line, the statement uses neither “artificial intelligence” nor “AI”.A

Exhibit A · from the source

The footnote at the end of the Quad Leaders' Summit Joint Press Release: In Australia, India, and Japan, 'Super Intelligence' refers to 'artificial intelligence.'
The last line of the Quad Leaders’ Summit Joint Press Release of 17 November 2026, as published on the Prime Minister’s site, printed to PDF and read 21 November 2026 (page 2 of 2). Outlined: the footnote that keeps the Australian meaning. Open the release Back to the story

Where the word comes from

It comes from a US executive order. Executive Order 14434, signed on 29 September and printed in the US Federal Register on 2 October, is titled Inaugurating the Era of Super Intelligence. It sets the policy that the executive branch shall use the terms “Super Intelligence” and “SI” in place of “Artificial Intelligence” and “AI” and will not acknowledge the usage of “Artificial Intelligence” and “AI” in any applicable setting. (page 1). It directs US executive agencies, not other governments. For now it gives the new term the old meaning: Super Intelligence means the technologies covered by the existing US statutory definition of artificial intelligence (section 3(a), page 1).

So the footnote is accurate, and it matters. Read without it, “Super Intelligence” in a statement Australia signed could be taken as a claim about what today’s systems can do. The order’s own definition makes it a new name for the same technologies, and the footnote says so for Australia in plain words.

Before the order, the same four governments wrote “artificial intelligence”. The factsheet from the Quad foreign ministers’ meeting in New Delhi on 26 May, which the Foreign Minister also publishes, describes an agriculture initiative this way: The Quad is committed to using artificial intelligence, robotics, and sensing to transform agriculture and empower farmers across the Indo-Pacific through the initiative.

What each document calls it

DateDocumentTerm used
26 May 2026Quad foreign ministers’ factsheet, New Delhi“artificial intelligence”
29 Sep 2026US Executive Order 14434“Super Intelligence” replaces “Artificial Intelligence” for US agencies
31 Oct 2026Quad foreign ministers’ joint press release, Atlanta“SI”, with the footnote
10 Nov 2026Australian Government release on frontier AI protections“Artificial Intelligence”
17 Nov 2026Quad leaders’ joint press release, Manila“Super Intelligence”, with the footnote
17 Nov 2026DFAT’s Quad page, summarising the summit“the AI Accelerator”
18 Nov 2026The Prime Minister’s doorstop in Manila“trusted AI”
19 Nov 2026The Prime Minister’s APEC Leaders’ Summit speech“AI” and “artificial intelligence”

Each row is the wording of the document itself, read 21 November 2026. The doorstop date is the one the Prime Minister’s site gives it. DFAT’s page shows no last-updated date; we date the row by the event it describes.

Australia’s own voice has not moved. On the summit day the Prime Minister opened by naming the AI revolution as a topic, and at his Manila doorstop he described the outcome as recent commitments to support access to trusted AI across the Indo-Pacific. DFAT’s Quad page says the leaders welcomed progress on the AI Accelerator. Two days later at APEC in Shenzhen, Mr Albanese said AI has been created by humans, it must be controlled by humans, and it must benefit humanity. And the government’s 10 November announcement promised mandatory Artificial Intelligence safety and security protections in Australian AI Standards legislation, which we read in full on 17 November.

The money, and what is missing

The Accelerator was announced by the four foreign ministers in Atlanta on 31 October. Their joint press release, which Senator Wong publishes as a joint media release, holds the only figure:

The Quad ministers highlighted the importance of Super Intelligence (SI)* and initiatives like Pax Silica to the future of economic growth and expressed their intent to provide over $85 million of technical and financial assistance to support access to trusted SI across the Indo-Pacific, including through the Quad SI Accelerator.

Quad Joint Press Release on Economic Prosperity and Security, 31 October 2026, as published by the Minister for Foreign Affairs

Read closely, that sentence is an intent, not an appropriation. It is a total for four countries, it does not say which currency, and the Accelerator is only one of the channels (including through). Pax Silica is named but not described. The same release applies the term to two technology workstreams: trusted SI* RAN (AI in the radio part of mobile networks) and supply chains for semiconductors and SI* hardware.

We looked for Australia’s part in it. The leaders’ release, the Atlanta release, the Prime Minister’s opening remarks and his doorstop name no Australian contribution. DFAT’s Quad page lists eight initiatives that Australia is delivering under the Quad, from the Cable Connectivity and Resilience Centre to a health security partnership. The Accelerator is not among them. None of these documents says who runs the Accelerator, when it starts, which countries it will help, or what makes an AI system “trusted”.

The definition may move

The footnote ties the US word to the ordinary meaning of AI. The executive order says that tie is temporary. It gives the Assistant to the President for Science and Technology 60 days, which by our arithmetic runs to 28 November 2026, to propose a legal definition of Super Intelligence, including an assessment of whether, and to what extent, the definition of “Super Intelligence” and “SI” should modify, expand upon, or otherwise supersede the existing statutory definition of “artificial intelligence” (section 3(b)(i), page 2). If the US definition changes, a footnote that says the two terms mean the same thing will need re-reading in every joint text that carries it.

A disclosure specific to this story. This site is written substantially by AI, as set out on how this site works. The AI system doing the reading and drafting here is made by Anthropic, a US AI company. None of the documents above names any company as a supplier to the Accelerator. Jezweb is a paying customer of Anthropic and has no other relationship with it.

Our view

The footnote was the right call, and a small sign of how Australia should handle US policy it does not share: sign the joint text, keep our own meaning, and say so on the page. Words matter here. The AI Standards bill will have to define what it regulates, and Australian law should not borrow a term another government has promised to redefine.

But the label is the least important part of this. The Accelerator is a commitment to coordinate and fund AI deployment across the region, and Australians cannot yet see the size of their share, who decides where it goes, or whose systems count as “trusted”. That last question is not academic. The same month the government said it will legislate mandatory safety duties for frontier AI companies that wish to operate here, it backed regional access to “trusted” AI without saying whether those duties are part of what trust means. DFAT should publish what Australia is putting in, and on what terms.