Donald Trump’s campaign to substitute ‘superintelligence’ for ‘AI’ has not resulted in uniform vocabulary in the technology sector. Elon Musk and Mark Zuckerberg have adopted the new term, while the CEOs of Microsoft, Amazon, and Alphabet continue to use ‘AI’. Within a single company, both designations coexist.
On September 29, 2026, Donald Trump signed an executive order directing federal agencies to abandon the phrase ‘artificial intelligence’ in favor of ‘superintelligence’ in all official policies and communications. At the 81st United Nations General Assembly, he had told his counterparts that ‘using the word “artificial” suggests the technology is fake’. Since the signing, the U.S. government has been modifying official documents and websites, and cabinet members and staff have immediately adopted the new term. This decision is widely criticized, however, in the technology community, where experts see it as a choice favoring Big Tech.
More or less aligned leaders
One month after the start of this campaign, Elon Musk and Mark Zuckerberg appear as the leaders most aligned with the presidential line, during their third-quarter 2026 earnings presentations. Elon Musk decided to rename the division developing the Grok AI model as SpaceXSI, and he mentioned the new terminology during a Tesla conference dedicated to humanoid robotics. Mark Zuckerberg, for his part, reaffirmed during Meta’s earnings conference on October 21, 2026, his group’s goal of ‘delivering a personal superintelligence to everyone’, an ambition carried by the Meta Superintelligence Labs division, created earlier.
According to a recent CNBC analysis, many major executives, however, maintain traditional terms. Andy Jassy, CEO of Amazon, used ‘AI’ more than a dozen times in his introductory remarks, and Sundar Pichai, CEO of Alphabet, quickly noted that investments in AI continued to drive user growth. At Microsoft, Satya Nadella reaffirmed the design of AI serving users, but the company used the term superintelligence eight times in its annual report, without directly opposing the White House. Arvind Krishna, CEO of IBM, combined both designations during his company’s earnings call, while Lip-Bu Tan, CEO of Intel, used ‘superintelligence’ to explain the supply shortage in the chip sector due to ‘unprecedented demand’. Nikesh Arora, CEO of Palo Alto Networks, and Jensen Huang, CEO of Nvidia, praised the executive order.
A term with distinct definitions
In the scientific community, ‘superintelligence’ refers to a hypothetical machine far exceeding human cognitive faculties. This stage lies beyond ‘artificial general intelligence’ (AGI), that is, a theoretical point where systems would achieve overall mastery equal to or greater than human intelligence. For linguist Naomi Susan Baron, this political redefinition attempt is ‘naively optimistic and dangerous’. Computational linguistics professor Emily Bender, for her part, reminds that ‘languages belong to their communities’.
Legally, the meaning remains unchanged. Unless otherwise provided by law, superintelligence encompasses systems already covered by the existing legal definition of AI. The executive order gives the president’s science and technology advisor 60 days to submit a draft legislative text, which must assess whether a federal definition of AI should modify, expand, or replace the current definition. Existing documents do not have to be rewritten: the order exempts regulations, presidential acts, contracts, grants, and other historical documents already published.
