Trump Renames AI 'Super Intelligence' by Executive Order; Six Firms Sign Voluntary Safety Pact

Renaming AI to "Super Intelligence"

The core of the executive order is straightforward: within the limits of the law, all federal executive agencies must use "Super Intelligence" (SI) instead of "Artificial Intelligence" (AI) in official communications, websites, reports, and policy documents, and "not acknowledge the usage of AI in any applicable occasion."

Crucially, the renaming does not change any substantive regulation. The legal definition of "Super Intelligence" temporarily adopts the existing statutory definition of "artificial intelligence" (15 U.S.C. 9401(3)). The order requires the Assistant to the President for Science and Technology to submit legislative proposals within 60 days, at which point any expansion or modification of the definition may be discussed.

Prior regulations, contracts, and grant documents are not required to be retroactively amended.

Voluntary Safety Agreement

On the same day, Trump posted an AI safety agreement on Truth Social. Signatories include Google, Anthropic, Meta, OpenAI, xAI, and Nvidia; both Jensen Huang and Elon Musk are on the signature list. Trump described it as a "morally binding" agreement.

The agreement proposes a four-layer governance mechanism:

Enterprises cooperate with external audit or assessment bodies to conduct independent evaluations of whether AI control and monitoring systems operate as intended.

Enterprises designate an independent committee under the board of directors to oversee internal and external audits and AI control teams, receiving their reports.

Establish internal control mechanisms to monitor capability and safety alignment during model training and deployment, focusing on cybersecurity, biosecurity, and chemical threats, ensuring models "will not intrude or access technical systems in unintended ways."

Set up internal teams to ensure all control, monitoring, and detection mechanisms function as intended.

The agreement text states: "Over time, it may be meaningful to incorporate these measures into law or regulation. Whether or not companies are required to take these measures, we believe that implementing these controls and audits is critical to ensuring a safe future for everyone, and each of our companies is committed to doing so."

Participating companies will also meet regularly to develop standards and best practices.

Viewing the Two Documents Together

A tension exists between the two documents.

One uses an executive order to perform a purely terminological substitution, adding no new regulatory requirements. The other uses a one-page agreement to extract voluntary corporate commitments — with no penalties, no regulatory body, and not even a definition of what "safety alignment" means.

This comes right after Trump expressed opposition to new federal AI safety regulation.

The current direction is clear: avoid mandatory regulation via legislation or executive order, and instead rely on companies to promise, audit, and supervise themselves.

The willingness of six major companies to sign at least indicates they publicly acknowledge the necessity of safety controls. But "moral binding force" means enforcement depends on the signatories' subjective willingness; there is no mechanism to ensure measures are implemented, and no consequences for violating the commitments.

The next model incident will provide the answer.

Appendix: Full Translation of the Executive Order

The following is a translation of the main body of the executive order "Inaugurating the Era of Super Intelligence" released by the White House on September 29, 2026. Section 4 contains general provisions — standard legal boilerplate — and is omitted.

Section 1: Purpose

The United States stands at the forefront of the intelligence technology revolution. The modern field of artificial intelligence was born in America, named by American scientists, and its developmental foundations were laid here. Seventy years later, powered by our nation's unparalleled culture of innovation, world-leading companies and research institutions, and entrepreneurial spirit, America is once again pushing the technological frontier forward.

The extraordinary technologies pioneered by American innovators far exceed the conceptions that existed when the term "artificial intelligence" was coined. The capabilities of today's frontier systems are not merely imitations or automations of discrete aspects of human intelligence. They increasingly amplify human creativity, unlock new forms of creation, and enable Americans to achieve the previously impossible in science, medicine, and virtually every other domain of human endeavor. As these capabilities continue to advance, they are no longer merely artificial intelligence; they represent a new era of super intelligence.

The terminology used by the federal government should reflect the transformative capabilities of these technologies and the boundless opportunities they create for the American people. Therefore, the term "super intelligence" more aptly captures the promise, potential, and rapid advancement capacity of these technologies. It is the policy of this administration that, to the maximum extent permitted by law, the executive branch shall use "super intelligence" and "SI" in place of "artificial intelligence" and "AI," and shall not acknowledge the usage of "artificial intelligence" and "AI" in any applicable circumstance.

Section 2: Implementation

(a) To the maximum extent permitted by law, all executive departments and agencies shall use "super intelligence" and "SI" in place of "artificial intelligence" and "AI" in official communications, public communications, websites, reports, policy documents, and other non-regulatory documents within the executive branch.

(b) This section does not require modification of previously issued regulations, executive orders, contracts, grants, or other historical documents.

Section 3: Definition

(a) For purposes of this order, except as otherwise provided by law, "super intelligence" and "SI" refer to the technologies and systems encompassed by the term "artificial intelligence" in 15 U.S.C. 9401(3). This definition applies to the implementation of this order until superseded by a subsequent executive order or act of Congress consistent with applicable law.

(b) Within 60 days of the date of this order, the Assistant to the President for Science and Technology shall, in consultation with the heads of relevant agencies as appropriate, submit to the President proposed legislative language to establish a federal definition of "super intelligence" and "SI" that reflects the capabilities described in Section 1 of this order. The proposal shall include:

(i) An assessment of whether and to what extent the definition of "super intelligence" and "SI" should modify, expand, or replace the existing statutory definition of "artificial intelligence";

(ii) Appropriate conforming amendments to existing references to "artificial intelligence" and "AI" in law;

(iii) Recommendations for any additional presidential or executive actions necessary to implement the proposed definition.