Trump Signs Executive Order Replacing 'Artificial Intelligence' with 'Super Intelligence'
The phrase "artificial intelligence" has appeared in federal policy documents since at least the 1980s. As of September 2026, it is no longer welcome in the executive branch's official vocabulary. A new executive order signed by President Donald Trump directs that official policy websites, policy documents, and press releases refer only to "Super Intelligence" — a rebranding of a technology that has shaped US regulatory strategy for years.
Trump framed the decision in characteristic terms. "The word super is the best word of all, and it's the simplest," he said, according to the order's announcement. The directive reportedly applies across the executive branch with no stated exemptions, though the summary available at publication did not detail which agencies or offices are subject to enforcement mechanisms.
The change follows a pattern of high-profile AI governance milestones in Washington. In October 2023, President Biden signed a sweeping executive order on AI that established safety testing requirements, reporting thresholds for large models, and coordinated agency guidance. That order, structured around the term "artificial intelligence," became the reference point for federal AI policy. Trump later rescinded it in early 2025 and replaced it with a deregulatory framework, a move that reshaped how agencies approached AI oversight. The new terminology order goes a step further: it changes not just policy direction, but the words agencies use to describe the technology itself.
The Trump Super Intelligence executive order does not exist in a vacuum. It arrives at a moment when the US, China, and the European Union are competing to set the definitions that shape international AI governance. Vocabulary, in that contest, is not a neutral detail.
How the Terminology Change Affects Federal Agencies
Federal agencies have spent years building internal vocabularies around AI. The Biden 2023 order required agencies to designate chief AI officers — more than two dozen positions were created across departments — and to publish inventories of AI use cases. Those inventories, by title and description, use the term "artificial intelligence." Under the new order, agencies will need to revise public-facing documents and policy pages to substitute "Super Intelligence."
Read next Laika's Wildwood: Stop-Motion Fantasy at TIFF 2026The scope of that revision is difficult to estimate from the available summary. The federal government operates hundreds of websites and publishes thousands of documents annually touching on computing and automation. NIST's AI Risk Management Framework, released in 2023 and updated since, is among the most widely cited federal AI documents. Its name uses the banned term. Whether documents of that kind will be renamed, reissued, or simply left with a note about the terminology shift is unclear.
Procurement is another pressure point. Federal contracts for machine learning tools often encode technical definitions in statements of work. If agencies must substitute terminology in solicitations, vendors and contracting officers will need to map "Super Intelligence" to existing technical categories. The General Services Administration and the Department of Defense, two of the largest federal buyers of AI systems, have not been described in the order's summary as receiving specific carve-outs.
Compliance costs could be modest. A 2023 study by the Brookings Institution estimated that federal agencies spent hundreds of millions of dollars annually on digital policy implementation, though terminology changes typically represent a small fraction of that. Still, the administrative layer matters: each renamed page and revised template consumes staff time and creates a paper trail.
What Is 'Super Intelligence' and Why Does the Label Matter?
"Super Intelligence" is not a standard technical term in the AI research community. The closest conventional phrase is "superintelligence," popularized by philosopher Nick Bostrom in his 2014 book of that name. In that literature, superintelligence describes a hypothetical system that surpasses human cognitive performance across nearly all domains — a far more ambitious concept than the large language models and machine learning systems currently deployed in government services.
That gap matters. Current federal AI applications include benefits eligibility screening, weather modeling, and cybersecurity threat detection. None of these are superintelligent by any accepted definition. Labeling them "Super Intelligence" creates a mismatch between the language of policy and the state of the technology.
Researchers have long debated whether terminology shapes outcomes. A 2022 survey of AI policy experts conducted by the Center for Security and Emerging Technology (CSET) at Georgetown found that framing affects public perception of risk and urgency. Terms that sound more powerful can inflate expectations and, in some cases, accelerate calls for regulation — or, conversely, justify deregulation on the grounds that benefits outweigh hypothetical dangers.
The label also carries international weight. The European Union's AI Act, which entered into force in 2024, classifies systems by risk tier — unacceptable, high, limited, and minimal — using explicit "artificial intelligence" definitions. China's Interim Measures for Generative AI Services, issued in 2023, and its broader algorithm registry rules use the term "人工智能" (artificial intelligence) throughout. If US agencies stop using the same vocabulary, joint standards work and treaty negotiations will require translation agreements that did not previously exist.
Reactions From the Tech Industry and Policy Experts
Most major technology trade associations have not commented publicly on the order in the summary available. That silence itself is notable. In previous AI policy debates — including the 2023 Biden order and the EU AI Act negotiations — industry groups such as the Information Technology Industry Council and the Chamber of Commerce's Technology Engagement Center issued detailed responses within days.
Policy analysts have been more vocal in preliminary assessments. Experts at Brookings and CSET have argued that terminology shifts in government can have measurable effects on funding cues, procurement language, and the behavior of agencies that take their signals from the White House. A term like "Super Intelligence" may signal to agencies that AI is a strategic priority, not merely an administrative tool. It may also signal the opposite: that the technology's perceived risks have been downgraded in official framing.
The scientific community has raised a separate concern. Researchers who publish in journals and present at conferences use "artificial intelligence" and its subfield names as standard. If US government documents use a different term, citations and cross-references between government reports and academic work become inconsistent. That friction is small per document, but it accumulates across a research ecosystem that depends on shared vocabulary.
Broader Implications for US AI Competitiveness and Global Standing
The US, China, and the EU together account for the overwhelming majority of global AI research spending and regulatory activity. A 2024 Stanford AI Index report estimated that global private AI investment exceeded $90 billion in 2023, with the US receiving the largest share. Coordination among these three blocs is already fragile. A terminology divergence adds another layer of complexity.
Diplomatic consequences are difficult to project. International standards bodies such as ISO/IEC JTC 1, which maintains AI terminology standards, operate by consensus. A US delegation that uses "Super Intelligence" while other members use "artificial intelligence" would need to negotiate a mapping, slowing work that already takes years.
There is also a signaling dimension. Alliances and rivalries in AI governance are partly constructed through shared language. If the US is seen as stepping outside that shared vocabulary, partners in the G7 and OECD may read it as a retreat from the consensus-based approach that characterized the Biden-era AI framework. Alternatively, some analysts argue that the shift could be interpreted as a domestic political move with limited international spillover, given that technical cooperation often continues below the level of official terminology.
What Comes Next: Enforcement, Compliance, and Open Questions
The executive order's implementation details remain the central unknown. The available summary does not specify a deadline for agency compliance, an enforcement mechanism, or a process for handling documents that cannot be easily renamed. Agencies may issue their own guidance in the coming weeks. NIST, which maintains the AI Risk Management Framework, and the Office of Management and Budget, which coordinates federal policy implementation, are the two bodies most likely to lead that work.
Legal questions also linger. Some federal documents — including statutes and prior executive orders — use "artificial intelligence" in their titles. An executive order cannot retroactively rename a law passed by Congress. Agencies will have to decide whether to reference those laws by their statutory titles or by paraphrased descriptions.
For companies, the practical effect may be limited in the short term. Federal contracts and grants are typically awarded against technical specifications, not terminology. But over a longer horizon, vocabulary shapes markets. If the US government consistently uses "Super Intelligence," vendors may adjust their own product naming and marketing to match.
The order's most lasting effect may be symbolic. For decades, "artificial intelligence" has been the shared language of policy, research, and industry. The Trump Super Intelligence executive order replaces that language — at least within one government — and asks the rest of the world to either follow, translate, or ignore. The answer to that question will not be settled in a press release. It will be settled in the slow work of procurement, standards, and diplomacy that follows.
Source: The Verge
