{"slug": "artificial-intelligence-has-a-new-name-the-white-house-should-make-sure-it-has", "title": "Artificial Intelligence Has a New Name. The White House Should Make Sure It Has the Same Oversight.", "summary": "The Department of Homeland Security has renamed its online AI use-case inventory the \"Super Intelligence Use Case Inventory\" and the FBI altered a 2019 National Defense Authorization Act quote to read \"any super system\" instead of \"any artificial system,\" following President Trump's September 29 executive order rebranding federal AI as \"Super Intelligence.\" The order requires White House Office of Science and Technology Policy director Michael Kratsios to send the President proposed legislative language defining \"Super Intelligence\" by November 28, and the article warns that a narrow definition tied to Section 1's frontier-system capabilities could put everyday government algorithms outside federal AI oversight. Senator Mark Warner (D-Va.), vice chairman of the Senate Intelligence Committee, said \"changing what we call artificial intelligence does nothing to address the very real risks that come with increasingly powerful AI systems.", "body_md": "The Department of Homeland Security now calls its [online list](https://www.dhs.gov/si/use-case-inventory) of artificial intelligence (AI) use cases the Super Intelligence Use Case Inventory. A disclaimer banner on the webpage explains that the term “AI” was changed where appropriate following President Trump’s September 29 [executive order](https://www.whitehouse.gov/presidential-actions/2026/09/inaugurating-the-era-of-super-intelligence/) (EO) rebranding the federal government’s AI usage as “Super Intelligence.” On the [FBI’s webpage](https://www.fbi.gov/investigate/counterintelligence/emerging-and-advanced-technology/super-intelligence) explaining “Super Intelligence,” the agency even altered a direct quote from the 2019 National Defense Authorization Act so that it reads “any super system,” even though the statute actually says “[any artificial system](https://www.govinfo.gov/content/pkg/PLAW-115publ232/html/PLAW-115publ232.htm).” So far, the rebranding and name change has altered nothing in law, but the White House science office owes the President a proposed “Super Intelligence” definition by November 28, and beforehand, it should clarify which laws will change.\n\nSection 3 of the order, “Definition,” has the potential to change a lot in the law. Under that provision, Michael Kratsios, the President’s science adviser and director of the White House Office of Science and Technology Policy (OSTP), is required to send the President [proposed legislative language](https://www.govinfo.gov/content/pkg/FR-2026-10-02/html/2026-20321.htm) for a federal definition of “Super Intelligence” (SI). As per the EO, this definition needs to reflect “the capabilities described in section 1,” which are that frontier systems “do much more than imitate or automate discrete aspects of human intelligence.” If that high bar replaces the broad definitions of AI in current law, everyday government algorithms could quietly fall outside federal AI oversight.\n\nBefore OSTP sends the proposed definition to the President, it should do two things:\n\n1.) It should publish a coverage table showing how the new definition would affect every law, memo, regulation and standard contract clause that currently uses the term “artificial intelligence.”\n\n2.) It should ensure existing oversight stays in place. No law, memo or contract clause should cover fewer federal systems than it does currently unless OSTP names it and provides a public justification answering why. Two moves can preserve current oversight: create a separate category for true frontier systems, and include a savings clause that keeps existing AI rules in place.\n\nCritics of the “super intelligence” EO have mostly characterized it as a branding exercise with no real implications on broader AI risks. Senator Mark Warner (D-Va.), vice chairman of the Senate Intelligence Committee, [responded](https://www.warner.senate.gov/newsroom/press-releases/warner-statement-on-president-trumps-executive-order-to-rebrand-artificial-intelligence/) to the EO by stating “changing what we call artificial intelligence does nothing to address the very real risks that come with increasingly powerful AI systems.” This is a fair point, as far as the label is concerned. The EO only [changes](https://www.arnoldporter.com/en/perspectives/blogs/enforcement-edge/2026/10/whats-in-a-name) the statutory scope of AI if it has congressional action behind it.\n\nThe upcoming definition change is a different story because the September EO contained no clause mandating OSTP write a narrow definition. In fact, when the Treasury Department [proposed new rules](https://www.federalregister.gov/documents/2026/10/26/2026-21786/outbound-investment-security-program-regulations) in October restricting U.S. investment in China (and other countries of concern), it defined “SI system” broadly enough to include conventional AI. But if OSTP instead drafts a narrow definition in line with Section 1, everyday AI tools at agencies could lose oversight.\n\nConsider how this could play out at an airport. At many checkpoints, the Transportation Security Administration’s (TSA) CAT-2 system compares a traveler’s face to the photo on their ID and gives the TSA employee a recommendation on whether they match. DHS’s [SI Use Case Inventory](https://www.dhs.gov/publication/si-use-case-inventory-library) lists CAT-2 as a deployed “high-impact” system, and its [FAQ](https://www.dhs.gov/si/use-case-inventory) confirms that facial recognition and capture counts as a subset of Super Intelligence. Under OMB memo [M-25-21](https://www.whitehouse.gov/wp-content/uploads/2025/02/M-25-21-Accelerating-Federal-Use-of-AI-through-Innovation-Governance-and-Public-Trust.pdf), high-impact AI use cases like CAT-2 require pre-deployment testing, an impact assessment, ongoing monitoring, and an appeal process. Under today’s rules, if an agency wants to exclude a high-impact system like CAT-2 from additional safeguards, it needs a formal waiver or public determination stating why the system is not high-impact.\n\nThose rules apply because M-25-21 gets its definition of AI from [Section 238(g)](https://www.govinfo.gov/content/pkg/PLAW-115publ232/html/PLAW-115publ232.htm) of the 2019 National Defense Authorization Act which covers machine-learning tools “designed to approximate a cognitive task,” like face-matching. Matching a face to an ID is the textbook example of automating a discrete task, not a frontier system as described in section 1 of the EO. If OSTP proposes a narrow definition of SI focused on frontier capabilities, and Congress updates existing AI laws to match it, TSA’s CAT-2 system could stop counting as a covered SI use case.\n\nThe task is harder because Section 3 of the EO calls for an assessment of whether the new definition should change, expand on, or supersede the “existing [statutory definition](https://www.govinfo.gov/content/pkg/FR-2026-10-02/html/2026-20321.htm)” of AI, along with “conforming amendments” to “existing statutory references.” Notice that the order refers to “the existing statutory definition” in the singular, yet federal law actually depends on two different definitions of AI, written for different purposes. Civilian research laws, such as the [CHIPS and Science Act](https://www.law.cornell.edu/uscode/text/42/18901), use the definition in [15 U.S.C. § 9401](https://www.law.cornell.edu/uscode/text/15/9401) (which covers machine-based systems that make predictions, recommendations or decisions for human-defined objectives). By contrast, laws governing how federal agencies use AI themselves, including the [Advancing American AI Act](https://www.law.cornell.edu/uscode/text/40/11301), rely on [Section 238(g)](https://www.govinfo.gov/content/pkg/PLAW-115publ232/html/PLAW-115publ232.htm) of the fiscal 2019 National Defense Authorization Act (which also explicitly covers machine-learning tools that approximate a single cognitive task). Agencies implementing the Trump administration’s AI-to-SI rename are already split**.** In October, a [National Institutes of Health grant notice](https://grants.nih.gov/grants/guide/notice-files/NOT-OD-27-011.html) defined “Super Intelligence” using the research statute definition. Meanwhile [the FBI’s website](https://www.fbi.gov/investigate/counterintelligence/emerging-and-advanced-technology/super-intelligence) uses the defense statute. Because the wording of the two statutes is different, a single replacement definition is likely to narrow at least one of the sets of laws unless the new definition is broad enough to encompass what both existing definitions currently cover.\n\nProviding a public coverage table by the November 28th deadline, or by the time the definition goes to Congress, would clarify where the new definition falls. The table would show which definition each statute, memo, regulation, and standard contract clause currently uses and explain if the proposed definition keeps coverage the same, expands it, or narrows it. OSTP should already have most of this information as a result of having to draft all the conforming amendments for the proposed definition.\n\nNext, OSTP should create a separate tier within “Super Intelligence” that includes “frontier systems.” The Trump administration already regulates AI this way, with duties on “high-impact” deployments, and models [sorted by capability](https://www.govinfo.gov/content/pkg/FR-2026-06-05/html/2026-11415.htm), using a classified benchmark to designate frontier models.\n\nBipartisan legislative language for an upper tier also already exists. A [2025 bill](https://www.govinfo.gov/content/pkg/BILLS-119s2938is/xhtml/BILLS-119s2938is.html) from Sens. Josh Hawley (R-Mo.) and Richard Blumenthal (D-Conn.) reserves the term “artificial superintelligence” for AI that can, among other traits, “match or exceed human cognitive performance and capabilities across most domains or tasks.” Applying this higher threshold to all federal AI laws would exempt tools like TSA’s face-matching system, while applying it just to an upper tier maintains routine system coverage but reserves heavy-duty rules only for frontier models.\n\nFinally, OSTP should include a savings clause that keeps every system currently covered by federal AI rules still covered under those rules. [Section 2(b)](https://www.whitehouse.gov/presidential-actions/2026/09/inaugurating-the-era-of-super-intelligence/) of the EO states that agencies do not have to go back and rewrite any regulations, contracts or grants they previously issued, just to swap out the word “AI.” That spares agencies from having to relabel old documents but it does not preserve existing coverage and Section 3 of the EO has no savings clause.\n\nThe Trump administration can rename AI to “Super Intelligence” if it wants. The important question is not the branding, but whether OSTP will show which laws will change before going beyond a simple rebrand and altering the definition of AI. To help answer this question, OSTP should publish the coverage table and keep existing oversight where it stands today.\n\n**Disclaimer:** Orion Policy Institute (OPI) is an independent, non-profit, tax-exempt think tank focusing on a broad range of issues at the local, national, and global levels. OPI does not take institutional policy positions. Accordingly, all views, positions, and conclusions represented herein should be understood to be solely those of the author(s) and do not necessarily reflect the views of OPI.", "url": "https://wpnews.pro/news/artificial-intelligence-has-a-new-name-the-white-house-should-make-sure-it-has", "canonical_source": "https://orionpolicy.org/artificial-intelligence-has-a-new-name-the-white-house-should-make-sure-it-has-the-same-oversight/", "published_at": "2026-11-06 18:17:03+00:00", "updated_at": "2026-11-06 18:22:06.760870+00:00", "lang": "en", "topics": ["ai-policy", "artificial-intelligence"], "entities": ["Department of Homeland Security", "Federal Bureau of Investigation", "White House Office of Science and Technology Policy", "Michael Kratsios", "Donald Trump", "Mark Warner", "Department of the Treasury"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/artificial-intelligence-has-a-new-name-the-white-house-should-make-sure-it-has", "markdown": "https://wpnews.pro/news/artificial-intelligence-has-a-new-name-the-white-house-should-make-sure-it-has.md", "text": "https://wpnews.pro/news/artificial-intelligence-has-a-new-name-the-white-house-should-make-sure-it-has.txt", "jsonld": "https://wpnews.pro/news/artificial-intelligence-has-a-new-name-the-white-house-should-make-sure-it-has.jsonld"}}