President Donald Trump’s executive order mandating the use of "Super Intelligence" (SI) in place of "artificial intelligence" in federal communications has triggered a rapid shift in terminology among key technology figures, though it faces significant linguistic and legal hurdles. The move, which brands the technology as "super" while casting "artificial" as a pejorative, has seen major industry players sign on, but it conflicts with existing statutory language and long-standing technical definitions.
What Happened
Following a United Nations speech in which Trump declared that "the use of the word artificial makes intelligence fake," he issued an executive order last week forbidding the use of "AI" wherever possible in federal contexts. The rebranding has been swift: the US Center for AI Standards and Innovation (CAISI) has been renamed the Center for Advancing Innovation and Standards for Super Intelligence (CAISSI), and Justice Department releases now describe fraud cases involving generated media as "Super Intelligence-Assisted Music Streaming Fraud." On Thursday, the White House further escalated the rhetoric, stating that anyone using the term "Artificial Intelligence" instead of "Super Intelligence" is "THE ENEMY."
Prominent figures in the SI industry have largely embraced the change. Elon Musk, referred to by Trump as his "First Buddy," promised to rename SpaceXAI to SpaceXSI. Jeff Bezos stated he supports the term because "artificial is not a very flattering thing," while Nvidia CEO Jensen Huang and OpenAI CEO Sam Altman have also adopted the terminology, with Altman noting that OpenAI itself will not change its name. Trump has also begun using "artificial" as a negative descriptor for other entities, recently labeling the press "artificial media" and dismissing concerns about war-driven price increases as "just artificial."
Why It Matters
The mandatory adoption of "SI" creates a disconnect between federal terminology and established technical and legal frameworks. The term "superintelligence" (one word) has existed since at least Nick Bostrom’s 2014 book and specifically refers to hypothetical systems that surpass human cognitive abilities, distinct from the current category of SI tools. By applying "SI" to everything from music generation tools to future machine intelligence, the administration risks diluting the term’s technical meaning. Furthermore, the rebrand undercuts industry efforts to market new, high-capability models as "superintelligent" if the term is already applied to ubiquitous, lower-tier systems.
Legal and legislative challenges also persist. Existing laws, such as the Take It Down Act, explicitly reference "artificial intelligence," requiring federal agencies to cite these statutes in court arguments regardless of the new executive order. Congress cannot be barred from using the original term, and congressional Republicans face the dilemma of adopting a label that is perceived as unintuitive and politically freighted with anti-regulatory sentiment. Conversely, pro-regulation figures like Senator Bernie Sanders have already used the traditional definition of "superintelligence" in bills like the Ban Artificial Superintelligence Act to advocate for restrictions on powerful systems, highlighting the term’s contested usage across the political spectrum.
The Bottom Line
While Trump’s linguistic campaign has secured quick adoption from high-profile SI leaders, the rebrand faces structural resistance from existing laws, technical definitions, and public sentiment. The administration’s attempt to elevate the technology’s image by renaming it "Super Intelligence" has not alleviated concerns about data centers or industry regulation, and the disconnect between federal vocabulary and statutory language creates ongoing compliance and communication challenges for the industry.