Super Intelligence (SI) Glossary: AI and SI Terms Defined
Updated October 11, 2026
Super Intelligence (SI), formerly called artificial intelligence (AI), is the name Executive Order 14434 directed federal executive agencies to use from September 29, 2026, and the name Iternal supports. Three communities use overlapping words for different things: the federal government's new name for the AI capability in use today; AI researchers' decades-old “superintelligence” (ASI), a still-hypothetical capability tier above human intelligence; and iternal.ai's own long-standing shorthand for its systems-integrator partners.
SI, AI, EO 14434, AGI, ASI and the 2026 Events
The same words now carry different meanings depending on who coined them and when.
- SI — Super Intelligence
- Super Intelligence, abbreviated SI, is the term U.S. federal executive agencies have been directed to use instead of "artificial intelligence" since Executive Order 14434 (Source: Federal Register, FR Doc. 2026-20321), signed September 29, 2026. The order defines SI by direct cross-reference to the existing statutory AI definition at 15 U.S.C. §9401(3) (Source: Cornell Law School Legal Information Institute) — SI names the same AI that already exists and ships today. The directive binds only the executive branch's own non-statutory communications "to the maximum extent permitted by law"; it does not change any existing law, contract, or regulation. Iternal backs the name because it describes measured results, not imitation: in a controlled trial of 95 professional developers building an HTTP server, those using GitHub Copilot finished 55.8% faster than the control group (Source: Peng, Kalliamvakou, Cihon and Demirer, arXiv:2302.06590). For more information visit the Why Iternal Supports Super Intelligence page.
- AI — Artificial Intelligence
- Artificial intelligence, or AI, is the established term for computer systems that perform tasks associated with human intelligence, and it remains the primary term in search demand, everyday usage, and the bulk of existing law. The U.S. government's own working definition — used by the National AI Initiative Act of 2020 and cross-referenced by Executive Order 14434 — defines AI at 15 U.S.C. §9401(3) as "a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments." (Source: Cornell Law School Legal Information Institute) Merriam-Webster defines it more simply as the capability of computer systems to imitate intelligent human behavior (Source: Merriam-Webster).
- Executive Order 14434
- Executive Order 14434, "Inaugurating the Era of Super Intelligence," is the September 29, 2026 presidential order that directed federal executive agencies to write "Super Intelligence" and "SI" in place of "Artificial Intelligence" and "AI" in official, non-statutory communications (Source: Federal Register, FR Doc. 2026-20321). Published in the Federal Register October 2, 2026, it defines the new terms by cross-reference to the existing AI statute rather than creating a new one, and it leaves existing regulations, contracts, and grants untouched. For the full walkthrough, see Executive Order 14434, explained for businesses.
- AGI — Artificial General Intelligence
- AGI describes a hypothetical AI system with roughly human-level breadth of cognitive ability across most domains — not a system that merely performs one task well. Physicist Mark Gubrud coined the term in 1997 (Source: Wired, "The Man Who Invented AGI"); OpenAI's 2018 charter defines it as "highly autonomous systems that outperform humans at most economically valuable work" (Source: OpenAI Charter), and Google DeepMind's 2023 "Levels of AGI" paper scores progress toward it on a six-level scale (Source: arXiv:2311.02462). No AI system has been generally agreed to have reached AGI as of this writing. See SI vs. AGI vs. ASI: what's the difference?
- ASI — Artificial Superintelligence
- ASI is the AI-research term for a still-hypothetical tier of capability above AGI: a system that substantially exceeds human-level general intelligence across virtually every domain, not merely matches it. Philosopher Nick Bostrom popularized the modern definition in his 2014 book Superintelligence: Paths, Dangers, Strategies, describing it as "any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest" (Source: Wikipedia, "Superintelligence"). ASI should never be confused with the federal government's "Super Intelligence (SI)," which names the AI capability in use today, not this hypothetical tier — see SI vs. AGI vs. ASI: what's the difference?
- Safe Superintelligence Inc. (SSI)
- Safe Superintelligence Inc. is an AI research company founded June 19, 2024 in Palo Alto and Tel Aviv by OpenAI co-founder Ilya Sutskever, Daniel Gross, and Daniel Levy, with the stated goal of building safe artificial superintelligence as "our mission, our name, and our entire product roadmap" (Source: ssi.inc). SSI has shipped no intermediate products and raised a $2 billion Series A in April 2025 at a $32 billion valuation (Source: Wikipedia, "Safe Superintelligence Inc."). Its use of "superintelligence" refers to the decades-old AI-research capability tier above AGI, not the federal government's 2026 name for the AI capability in use today.
- White House Accord on Super Intelligence
- The White House Accord on Super Intelligence is a voluntary safety-governance pact signed September 29, 2026 — the same day as Executive Order 14434 — by the White House and six technology leaders: Sundar Pichai (Google), Dario Amodei (Anthropic), Mark Zuckerberg (Meta), Greg Brockman (OpenAI), Jensen Huang (NVIDIA) and Elon Musk (Source: CIO.com). It commits signatories to internal safety monitoring, independent audits, and board-level oversight of frontier models. It is not legally binding, and its own operative text refers to "frontier models," not "Super Intelligence," outside its title — signing it is not a commitment by any of the six companies to use the term "SI" in their own products or communications (Source: Al Jazeera).
- Ban Artificial Superintelligence Act
- The Ban Artificial Superintelligence Act of 2026 (H.R. 10538) is federal legislation from Senator Bernie Sanders and Representative Greg Casar, announced September 3, 2026 (Source: Unite.AI). The sponsors' summary defines “Artificial Superintelligence” as a system that “exhibits or can easily be modified to exhibit capabilities that match or exceed human cognitive performance... across a broad range of domains,” or one with “sufficient capabilities to plan and execute the disempowerment of humanity” (Source: Sen. Sanders' office, bill release summary) — the hypothetical ASI capability tier, not the federal government's “Super Intelligence (SI),” which names the AI capability in use today.
When “SI” Means Something Else Entirely
There is no single SI meaning: the same two letters carry several unrelated definitions, including one that iternal.ai itself has used for years, well before the federal order existed.
- SI — Systems Integrator (on iternal.ai)
- On iternal.ai, "SI" has historically meant systems integrator: the partner firms that deploy Iternal's software for enterprise and federal clients. This is a different meaning from the federal "Super Intelligence (SI)" defined above, and the two should never be confused. For Iternal's own government-buyer and partner-channel content, see AI for Federal Contractors.
- SI — NIST 800-53 Control Family
- In U.S. federal compliance documentation, "SI" also abbreviates System and Information Integrity, one of the control families in the NIST Special Publication 800-53 Revision 5 catalog used for federal authorization to operate, covering controls such as flaw remediation, malicious-code protection, and system monitoring (Source: FedRAMP control catalog). This is unrelated to both the federal "Super Intelligence (SI)" label and the systems-integrator usage above; context — a FedRAMP control list, a federal communications policy, or a partner reference — is the only reliable way to tell the three apart.
Quick Disambiguation Table
The same two letters carry at least six unrelated meanings; the surrounding context almost always tells you which one is meant.
| Sense of “SI” | What it means | How to tell |
|---|---|---|
| Super Intelligence | The federal government's 2026 term for AI (Executive Order 14434) | Appears in federal-agency communications or writing about the order itself |
| Systems integrator | An enterprise or channel partner firm that deploys software for a client | Appears in partner, channel, or case-study copy ("SI partnerships," "VAR/SI") |
| SI control family | NIST 800-53’s "System and Information Integrity" control family | Appears in a compliance or ATO control list alongside codes like AC, AU, CM, SC |
| SI units | The International System of Units (meters, kilograms, seconds) | Appears in scientific, engineering, or measurement contexts |
| Sports Illustrated | The U.S. sports and media magazine brand | Appears in sports, media, or publishing contexts |
| Spanish "sí" | The word "yes" in Spanish | Appears in Spanish-language text, usually with the accent mark |
FAQ: Super Intelligence (SI) Glossary
SI has several meanings. Since Executive Order 14434 (signed September 29, 2026), U.S. federal agencies use “SI” for “Super Intelligence,” their new name for AI. On iternal.ai and across the technology industry, “SI” has also long meant “systems integrator”; in federal compliance documents it can mean the NIST “System and Information Integrity” control family; elsewhere it means SI units. Context determines which meaning applies.
No. Historically on iternal.ai, “SI” referred to systems integrator — the partner firms that deploy Iternal's software for enterprise and federal clients. That usage is unrelated to the federal government's 2026 “Super Intelligence (SI)” label.
AGI (artificial general intelligence) describes a hypothetical AI system with roughly human-level breadth of ability across most domains; ASI (artificial superintelligence) describes a further, still-hypothetical tier that substantially exceeds human-level intelligence across virtually every domain. Neither has been built, and neither is the same as the federal government's “Super Intelligence (SI),” which names AI technology that already exists and ships today.
Safe Superintelligence Inc. (SSI) is an AI research company founded in June 2024 by OpenAI co-founder Ilya Sutskever, Daniel Gross, and Daniel Levy, with building safe artificial superintelligence as its sole stated goal. SSI's use of “superintelligence” refers to the AI-research capability tier above AGI, not the 2026 federal name for the AI capability in use today.
The White House Accord on Super Intelligence is a voluntary safety-governance pact six technology leaders — Sundar Pichai, Dario Amodei, Mark Zuckerberg, Greg Brockman, Jensen Huang and Elon Musk — signed with the White House on September 29, 2026. It commits signatories to internal and independent safety audits of frontier models; it contains no commitment by any signatory to use the term “Super Intelligence” or “SI” in its own products or communications.
This glossary covers the AI-to-SI terminology event specifically — SI, AGI, ASI, Executive Order 14434, and related disambiguation. For general AI and data terms such as LLM, RAG, tokens, SLM, and context window, see Iternal's AI glossary for LLM, RAG and token definitions.
Go Deeper on the AI-to-SI Transition
Executive Order 14434 changed federal vocabulary, not the technology. The hub covers what changed, what did not, and what it means for your organization.