• The Trump administration's push to replace 'AI' with 'Super Intelligence' in federal documents is underway, but implementation is uneven and the legal definition of AI remains unchanged.
  • The rebranding has extended to international diplomacy and substantive policy, with Treasury using 'superintelligence' in a recent investment restriction proposal.
  • A forthcoming legislative proposal on redefining AI could have more significant consequences than the terminology shift itself.

From AI to SI: A Rebradning with Limits

Federal agencies are gradually adopting the term "Super Intelligence" (SI) in official communications, following President Trump's September 29 executive order directing the replacement of "AI" and "Artificial Intelligence" with "SI" and "Super Intelligence" wherever legally permitted. The order applies to official correspondence, websites, reports, policy documents, and other non-statutory executive-branch materials, but explicitly excludes previously issued regulations, contracts, grants, and historical documents.

Implementation has been uneven. A Politico review on October 21 found that several agencies continued to use "artificial intelligence" on their main webpages. The order sets no general deadline for completing the transition, and legal exceptions mean both terms will persist in government materials for the foreseeable future. The State Department has extended the change to new emails, speeches, websites, official titles, and program names, according to an internal cable reported on October 20.

Legal Definition Unchanged, For Now

Crucially, the order preserves the existing statutory definition of artificial intelligence. The new term "Super Intelligence" expressly incorporates that definition, meaning the rebranding does not imply that current systems have achieved capabilities beyond human intelligence. Technical experts distinguish superintelligence from today's generative AI systems, and the order does not alter that distinction.

A more consequential provision requires the president's science and technology adviser to submit proposed legislative language within 60 days—by approximately November 28—assessing whether a new definition should modify, expand, or supersede the existing statutory definition. That proposal, if pursued, could ultimately lead to legal changes, but it would not itself alter the statute.

International Push and Policy Implications

The administration is pressing international partners to adopt the terminology. Treasury has been advocating for "super intelligence" in negotiations over an International Monetary Fund steering-committee statement, though Reuters (TRI) reported resistance. On October 22, Treasury used "superintelligence" in announcing proposed restrictions on certain American high-technology investments in China and other countries of concern—a sign that the terminology is entering substantive policymaking. That investment rule remains a proposal, not a final rule.

The broader policy context includes a voluntary safety agreement signed on September 29 by Nvidia (NVDA), SpaceX (SPCX), OpenAI, Anthropic, Meta (META), and Google (GOOGL), which includes internal controls and independent external auditing but no stated consequences for noncompliance. The administration has emphasized innovation and competition with China while relying on voluntary safeguards rather than new comprehensive regulation.

Mixed Reactions and Unclear Economic Impact

For businesses, the immediate implications are primarily administrative: federal-facing communications may increasingly use "SI," while existing contracts and grants need not be rewritten. The directive announces no new corporate tax, spending program, or technology restriction, and no measurable economic effect from the renaming alone has been established.

Safety advocates question whether rebranding and voluntary commitments adequately address risk. A September Reuters/Ipsos (IPS.PA) poll found that three-quarters of Americans worried AI companies had not done enough to prevent serious societal harm. However, the expert debate is not entirely one-sided. The Council on Foreign Relations' Kat Duffy characterized the administration's approach as acknowledging concerns without abandoning its anti-regulatory stance, while Samuel Hammond of the Foundation for American Innovation saw the company agreement as a meaningful acknowledgment that safety and security warrant closer coordination.

What to Watch

The key near-term development is the forthcoming definition proposal from the president's science and technology adviser. Whether it recommends changes to the statutory definition of AI could determine if the rebranding has substantive legal consequences. Meanwhile, expect continued uneven adoption of "SI" across agencies and ongoing diplomatic friction over the terminology. The Cloud Security Alliance has identified the definition proposal—not the name substitution—as the element most likely to produce meaningful consequences.

Correction: An earlier version of this article misstated the deadline for the science adviser's proposal. It is 60 days from September 29, approximately November 28.