AI is now 'Super Intelligence'. What actually changed?

On 29 September 2026 President Trump signed Executive Order 14434, titled "Inaugurating the Era of Super Intelligence". We read the order and the White House fact sheet on whitehouse.gov. Here is what the text actually contains.

• The policy. Section 1 says today's frontier systems "increasingly represent not merely artificial intelligence, but a new era of Super Intelligence". It makes it the policy of the administration that, "to the maximum extent permitted by law", the executive branch uses "Super Intelligence" and "SI" in place of "Artificial Intelligence" and "AI". It adds that the executive branch "will not acknowledge the usage of 'Artificial Intelligence' and 'AI' in any applicable setting".
• What agencies must change. Section 2 applies to official correspondence, public communications, websites, reports, policy documents and other "non-statutory documents within the executive branch". Old regulations, presidential actions, contracts and grants do not need to be altered.
• What the words mean. Section 3 defines "Super Intelligence" and "SI" as the technologies and systems "encompassed by the term 'artificial intelligence' as defined in section 9401(3) of title 15" of the US Code. In other words, the legal definition is unchanged.
• A homework assignment. Within 60 days, the President's science adviser must send the President proposed legislative language for a federal definition of Super Intelligence, including whether it should replace the current statutory definition. Counting from 29 September, that is about 28 November 2026.
• No new rights. Section 4 says the order creates no right or benefit enforceable by anyone against the government.

The fact sheet gives the reason. It says the name "Super Intelligence" better conveys the capabilities being developed, which it says go beyond "imitating or automating human intelligence as is implied by the term 'Artificial Intelligence.'"

One more thing the text does not do: it does not say that today's systems are superintelligent in the technical sense that researchers use. For that meaning, and how it differs from AGI, see our guide to superintelligence vs AGI.

The accord announced the same day

News outlets reported that the White House also announced a voluntary industry accord. Nextgov and Al Jazeera report that executives from companies including Google, Anthropic, Meta, OpenAI, xAI and Nvidia took part. The reports describe layered oversight: internal controls, an internal oversight team, independent outside auditors and an independent committee. Al Jazeera reports that the accord is not legally binding and carries no penalty. Nextgov reports that Vice President Vance rejected an FDA-style or FAA-style regulator for frontier models and said existing agencies already have authority over consumer harms.

The name change sits on top of a policy direction that began in January 2025.

1. January 2025Executive Order 14179 states a policy "to sustain and enhance America's global AI dominance" and revokes the Biden administration's AI safety order, EO 14110.
2. July 2025America's AI Action Plan sets out three pillars: accelerate innovation, build infrastructure and lead in international AI diplomacy and security. Its introduction says "The United States is in a race to achieve global dominance in artificial intelligence". The plan does not use the words AGI or superintelligence, in the text we read.
3. December 2025Executive Order 14365 targets state AI laws. According to the White House fact sheet it creates a Justice Department AI litigation task force, and it quotes the President: "We MUST have one Federal Standard instead of a patchwork of 50 State Regulatory Regimes."
4. March 2026The White House released a legislative framework asking Congress to override "unduly burdensome" state laws, according to law-firm and think-tank summaries. The September fact sheet calls it a National Super Intelligence Policy Framework.
5. September 2026Executive Order 14434 and the voluntary accord.

It is worth noticing how the vocabulary moved. In 2025, administration officials described "AGI" and "superintelligence" talk as a distraction. MIT Technology Review quoted the White House AI adviser Sriram Krishnan saying that "this notion of imminent AGI has been a distraction and harmful and now effectively proven wrong". A year later, the administration itself adopted "Super Intelligence" as its official label. We report both facts side by side and leave the conclusions to you.

We cannot know anyone's private motives, so we separate what was said from what others infer.

What the administration says

• The fact sheet says the name should reflect "the limitless promise" of the technology, and that the United States leads the world in it.
• The wider agenda is stated in the Action Plan: winning a race. Its introduction says "The AI race is America's to win". PolitiFact quotes the President as saying that "whoever wins super intelligence wins", and describing the United States and China as the two players.
• At the United Nations, according to MeriTalk, the President's science adviser Michael Kratsios said the United States "totally rejects" a global control scheme for superintelligence and favors "permissionless innovation".

What analysts say

• It is a race and dominance strategy. Brookings and other think tanks read the chip-export loosening and the infrastructure push as part of a competition with China. Brookings' Mark MacCarthy argued in February 2026 that loosening export controls may be right if superintelligence is not imminent.
• It favors light national rules over state ones. Lawfare called the March framework "industry-friendly" and expected Congress to be unlikely to pass it as written. CSIS argued that targeting state laws "undermines rather than advances" US technology leadership.
• The label is branding. PolitiFact quoted outside experts who said a new name could mislead, or fail to reassure a public that polls show to be more concerned than excited. The Pew Research Center found in 2025 that 50% of US adults were more concerned than excited about AI in daily life.

None of these readings can be proven from the documents. Notice also that the people disagreeing are often disagreeing about the facts: whether superintelligence is near. Our guide to what the forecasts really say covers that.

• The federal definition, due around 28 NovemberThe science adviser's legislative proposal could widen or replace the statutory definition of AI. Definitions decide which products fall under which rules.
• State laws and the preemption fightFederal pressure on state laws is real, but Congress has not passed a law overriding them as of this writing, and state rules in places such as California and New York still apply. Plan for both.
• Government contract languageIf you sell to US agencies, expect the new vocabulary in solicitations. Check whether your contracts and compliance documents use terms that map cleanly to the statutory definition.
• Voluntary commitments you could mirrorInternal controls, an oversight owner, independent review and a challenge function are good practice at any scale. See our guide to running AI agents safely.
• Your other marketsEU and UK rules did not change on 29 September. See our guide to the EU AI Act for small teams.