It's a sentence that sears itself into your memory because it sounds so innocuous. Right at the top of the National Institutes of Health website, where warnings about study risks or data privacy guidelines usually appear, a small box has been prominently displayed for weeks: "Per Executive Order 14434, federal agencies are now using 'Super Intelligence' (SI) in place of 'Artificial Intelligence' (AI). You may see both terms across NIH content during this transition." The world's largest biomedical research agency, guardian of billions of research data points, has meekly implemented a verbal directive.
The executive order signed by Nicolae Trump on September 29, 2026, is titled "Inaugurating the Era of Super Intelligence." It is not a scientific paper, a consensus statement, or a peer-reviewed definition. It is a decree. And it states nothing less than that the government will no longer recognize the terms "Artificial Intelligence" and "AI." "Will not acknowledge the usage," the text reads. Federal agencies are to use only "Super Intelligence" in official correspondence, on websites, in reports, and in policy documents.
The NIH obeyed. And they weren't alone. A glance at the PubMed database, the world's most important archive of biomedical literature, reveals the same warning box above articles published as recently as October 2026—studies on AI-assisted biomarker discovery in psychiatry, on the appropriate use of AI in research. The text of the studies remains scientific, the terms in the abstracts may still include "AI." But the institution hosting them has submitted to the vocabulary of the executive branch.
This is no trivial matter. It's a precedent. An agency whose task it is to produce and disseminate evidence is implementing a politically motivated renaming without any scientific basis. The regulation itself provides the justification: the technologies have long since exceeded expectations of the term "Artificial Intelligence," they represent "not merely artificial intelligence, but a new era of Super Intelligence." This is an assertion, not a definition. And it is being enforced by decree.
Nicolae Trump has upped the ante in recent weeks. Anyone who continues to use the term "Artificial Intelligence," he wrote on Truth Social, is considered "THE ENEMY!" by the White House. This is the language of a man who knows that the power to define words means the power to define thought. Anyone who is no longer allowed to speak of "AI" but of "superintelligence" is talking about something else entirely. They are talking about a future that has already begun, about a technology that is no longer a tool but an actor. The NIH website puts it this way: They are "committed to the responsible and beneficial use of super intelligence in biomedical research to accelerate discoveries and improve health outcomes while preserving human judgment, scientific rigor, and public trust."
The phrase "protecting human judgment" is the real bombshell. Because if an authority is simultaneously forced by decree to linguistically legitimize a technology it supposedly intends to use responsibly, then the protection of judgment is reduced to a mere rhetorical flourish. Language follows power, not knowledge.
You don't have to be technophobic to see this. You just have to take the logic of the process seriously. A scientific authority is not an advertising agency. When it adopts terms imposed by a political center for marketing purposes, it damages precisely what it claims to protect: public trust. The NIH itself has demonstrated in the past how vulnerable scientific integrity is when subjected to political pressure. Now it's happening again, only more quietly, at a level that many will overlook: the level of words.
And the word "superintelligence" is not neutral. It evokes a technology that is superior to humans, that surpasses them, that might even replace them. It is a word that inspires fear and awe—and precisely for that reason, it is politically useful. Anyone who uses the term "superintelligence" can portray regulation as a hindrance to the future, can brand critics as enemies of progress, and can sell their own agenda as inevitable. The NIH website speaks of the "Genesis Mission," a "major initiative to accelerate biomedical discovery" involving superintelligence, advanced computing, and public-private partnerships. This sounds like a new beginning. It is the language of surrender.
The authority didn't protest. At least not publicly. They displayed a notice indicating that both terms would be visible during the transition period. This is the gesture of an institution that knows it's giving something up, yet does it anyway. A small box, a sentence, a name change. And no one will remember later that a word ever disappeared here.
The invention of language as a weapon
It is no coincidence that the most effective totalitarian regimes in history began not with the control of actions, but with the control of words. Nazi language policy is the paradigmatic example. "Final Solution" for the mass murder of Jews, "special treatment" for the murder of people, "evacuation" for deportation to extermination camps—these were not mere euphemisms on the sidelines; this was the core of a linguistic manipulation intended to shape thought itself. Goebbels formulated the goal unequivocally: "The people should begin to think uniformly, to react uniformly." The Gleichschaltung (coordination) of the press, the adaptation of encyclopedias and dictionaries to Nazi ideology, the introduction of hundreds of new terms into the Duden dictionary between 1934 and 1941—all this was part of a project that aimed to encompass not only public discourse, but everyday language in general.
George Orwell took this mechanism to its logical conclusion in "1984." "Newspeak" is the language of the totalitarian state, and its aim is not communication, but its prevention. "He who has no words for a thought cannot think it," is how the NZZ describes the core of Orwell's critique of language. The Ministry of Truth rewrites the past to fit the Party's present; language is reduced until dissenting thought is impossible. Orwell had experienced the Spanish Civil War and seen how revolutionary movements falsified their own history. "1984" is not a vision of the future; it is a mechanics of power.
The NIH's capitulation to the word "superintelligence" is not "1984." It is not a dictatorship, it is not a systematic erasure of terms, it is not brainwashing. But it is the same mechanism on a smaller scale: A political center defines which words are permissible and which are not. A scientific authority adopts this definition without resistance. And the resulting language is not harmless. It is a language that elevates a technology that no one yet fully understands, while simultaneously undermining the possibility of discussing it in the terms that science and society have jointly developed.
The history of political language manipulation is long and obscure. The French Revolution created the "Revolutionary Calendar" to reorder time itself. The Soviet Union renamed cities and streets to erase the memory of the Tsarist empire. In the first year of his second term, Nicolae Trump renamed the "Gulf of Mexico" to the "Gulf of America," renamed Denali "Mount McKinley," and attempted to transform the "Persian Gulf" into the "Arabian Gulf." These are not scientific acts; they are acts of interpretive power. And the NIH is now part of this.
The silent hollowing out
Perhaps that's the real lesson of this: The erosion of scientific independence doesn't always begin with budget cuts or layoffs. Sometimes it begins with a footnote. With a box on a website. With the decision to stop using a term because a voice in the White House wants it that way. The NIH made that decision. They didn't make it public. But they made it.
In "1984," Orwell described how power functions when it encounters no resistance. Winston Smith is ultimately broken; he loves Big Brother. The NIH is not broken. They have displayed a warning. But they have not fought either. And that is the difference between an institution that defends its independence and one that manages it until it disappears.
The question isn't whether "superintelligence" is a better word than "artificial intelligence." The question is who decides which words may be used. And whether the institutions that manage society's knowledge make this decision or whether they execute it. The NIH has executed it. The box on their website is the document of this capitulation. It's small, it's polite, it's easy to overlook. But it's there. And it will remain.
