From 'AI' to 'Super Intelligence': What Executive Order 14434 Means for Clinical Practice
A vendor sends an email. The product you have been using for eighteen months — the AI-assisted imaging tool cleared by the FDA in 2024 — has a new name in the subject line. Powered by Super Intelligence. The tool has not been updated. The algorithm has not changed. The FDA clearance number is the same. The performance data is the same. The false positive rate is the same.
The name is different.
That gap — between the political terminology and the clinical evidence — is not trivial. Language shapes how physicians and patients calibrate trust in a tool. And in clinical practice, miscalibrated trust is a safety issue.What actually happened
On September 29, 2026, the President signed Executive Order 14434, titled Inaugurating the Era of Super Intelligence. The order directs all federal agencies to replace the terms "artificial intelligence" and "AI" with "Super Intelligence" and "SI" in their websites, communications, and official materials.
NIH is currently mid-transition. As of October 2026, the NIH.gov web modernization FAQ explains: "You may see both terms used across NIH content. We appreciate your patience as we work through these updates."
The FDA, CMS, and other health-focused agencies are expected to follow.
This is a language change. It is not a new regulatory standard, a new clearance pathway, or a new reimbursement framework. No AI medical device has been retested. No clinical validation study has been redesigned. The 1,614 FDA-cleared AI-enabled medical devices — soon to be referred to as SI-enabled medical devices — will carry the same clearance dates, the same clinical indications, and the same published performance data they had before the executive order was signed.
The tool that was "AI" last month is "Super Intelligence" this month. The performance data is the same.
What "Super Intelligence" implies that "AI" did not
The phrase "artificial intelligence" carries its own cognitive baggage, but it also carries an implicit admission of limitation. Artificial. Simulated. A constructed approximation of human reasoning, bounded by its training data and validation study.
"Super Intelligence" implies something categorically different. Superhuman. Exceeding human capability. Beyond ordinary intelligence, not just replicating it.
When a clinical tool is labeled "Super Intelligence," the physician's brain — and the patient's brain — starts from a different calibration point. The unconscious assumption about what the tool can do has been upgraded. The threshold for disagreeing with its output has been raised. Neither of these changes is based on new evidence.This is not speculation about worst-case behavior. It is how language and framing consistently operate in clinical settings. Studies on how label changes affect prescribing, diagnostic confidence, and shared decision making have documented this pattern for decades. The mechanism is not unique to AI — it is a feature of how humans assign credibility to named entities.
Apply that mechanism to a tool whose actual performance ceiling has not changed, and the result is a widening gap between perceived capability and demonstrated capability. That gap is where clinical errors live.
The automation bias connection
Automation bias — the tendency to favor an automated system's output over competing information — forms faster when a tool is trusted more. The research is consistent on this point: trust accelerates the Habit Cycle, and the Habit Cycle eventually runs without deliberate scrutiny.
The rebrand from AI to Super Intelligence is, in effect, a trust accelerant. It does nothing to the numerator — the performance of the tool. It operates entirely on the denominator — the scrutiny the physician applies before accepting the recommendation.
A physician who would have paused to question an "AI-generated" recommendation may not pause at the same rate for a "Super Intelligence" recommendation. The algorithm has not changed. The pause has been designed out by a name.For a full account of how automation bias forms and what to do about it, see Automation Bias in Healthcare: The Hidden Risk of Trusting Clinical AI Too Well.
What remains unchanged — and what physicians must hold onto
The 7 Red Flags for clinical AI evaluation were developed to assess what the evidence actually shows about a specific tool. They are immune to rebranding, because they are not about labels. They are about data.
The same questions apply to every device regardless of what it is now called:
- Regulatory status: What specifically did the FDA clear this tool to do? "SI-enabled" is not a clearance indication. The clearance document is.
- Explainability: Does the output show which features drove the recommendation — or does it simply assert a conclusion?
- Training data demographics: Was the tool validated on patient populations comparable to yours?
- Physician-in-the-loop design: Does the workflow require physician interpretation before action — or does the tool route around it?
- Real-world vs. trial performance: What does the published validation study show, and what does the post-deployment data show? These are often different.
- Adverse event reporting: Is there a mechanism for reporting errors and monitoring outcomes after deployment?
- Failure disclosure: Does the vendor publish its failure modes and the conditions under which performance degrades?
The regulatory record has not caught up
As of late October 2026, the FDA's official clearance database is still titled "Artificial Intelligence-Enabled Medical Devices." The page still reads: "the FDA encourages the development of innovative, safe, and effective medical devices, including devices that incorporate artificial intelligence (AI)." Not Super Intelligence. Not SI.
NIH is visibly mid-transition — its web modernization FAQ explicitly acknowledges the executive order and notes that both terms will appear across NIH content during the changeover. The FDA has made no equivalent announcement. Its 1,614-device list, its clearance documentation, and its Digital Health and AI Glossary all still use "artificial intelligence."
That gap has a practical consequence. A vendor can update their product sheet to say "SI-powered" today — that is now federally correct terminology. But the FDA clearance that device holds still says "AI-enabled." The physician reading the vendor marketing and the physician reading the clearance document are reading two different language systems describing the same tool.
The clearance documentation is the legal and evidentiary foundation of how a device is supposed to be used. Right now, the political label has moved ahead of the regulatory record — and no one has yet explained what happens when the two conflict in a clinical or legal context.CMS has similarly made no announcement about reimbursement coding. The gap between cleared devices (~1,614) and reimbursed ones (~10) is unchanged. Renaming the category does not close it.
What to tell patients
Informed consent language at many institutions currently references "AI-assisted" diagnosis, decision support, or monitoring. As vendors and institutions update their materials to comply with the executive order, patients will begin seeing "Super Intelligence" in documents they are asked to sign.
That shift requires a clinical conversation, not just a document update.
The most useful framing for patients is specific: describe what the tool does, what it has been cleared to detect, how often it is right, and what role the physician plays in the final decision. A consent conversation built on those specifics is more meaningful — and more honest — than any label on the document.
The answer to "Is this Super Intelligence?" is: "It is a cleared medical device that assists with [specific function]. My role is to review its output and make the final clinical judgment. Here is what it looks good at, and here is where it has limits."
That conversation is unchanged by the executive order. That conversation was always the right one.
What to do now
When you see "SI-powered" in a vendor update: ask for the FDA clearance number and read the clinical indication. The clearance is what the tool was actually evaluated to do. The label is not.
When you review AI — now SI — recommendations in the EHR: your evaluation criteria do not change because the label changed. Apply the same scrutiny. Generate your own clinical impression before reading the output. Override when the data supports overriding.
When procurement conversations happen: ask how the vendor defines "Super Intelligence" as a performance standard and how that definition maps to the FDA clearance criteria. The answer will tell you whether the terminology shift reflects a substantive claim or a compliance update.
When patients ask: explain the tool by what it does and what the evidence shows, not by what it is called.
The terminology changed. The evidence did not. Your job is to keep those two things from diverging in the clinical decisions you make on behalf of patients.Frequently Asked Questions
Q: Does Executive Order 14434 change FDA clearance requirements for AI medical devices?
A: No. The executive order changes terminology — it directs federal agencies to use "Super Intelligence" in place of "artificial intelligence." It does not establish new clearance pathways, revise the 510(k) process, or require existing cleared devices to undergo additional validation. The performance standards and clearance criteria for medical AI devices remain unchanged as of October 2026. The FDA has not yet announced how it will update existing clearance documentation.
Q: Will "Super Intelligence" tools be held to higher performance standards than "AI" tools were?
A: Not based on anything in the executive order. The order is a language directive, not a regulatory standard. Whether the FDA or CMS use the rebrand as an opportunity to raise performance requirements for a new category of SI-enabled devices remains to be seen — as of this writing, no such proposal has been issued. The current gap between the label and the evidence base is exactly why physicians should be watching this space closely.
Q: How should physicians handle the terminology change in clinical conversations with patients?
A: Ignore the new label in the consent conversation and focus on substance. Describe what the specific tool does, what FDA indication it holds, what its published performance looks like in populations comparable to your patient, and what your role is in the final decision. Patients who understand what a tool actually does are better positioned to give meaningful informed consent than patients who have been told the tool is "Super Intelligence."
Q: Does the rebrand affect CMS reimbursement for AI-enabled medical devices?
A: Not yet. CMS has not announced reimbursement changes tied to the executive order. The reimbursement gap — approximately 10 codes for 1,614+ cleared devices — remains. Whether the rebrand to "Super Intelligence" creates political pressure to expand reimbursement is an open question. If anything, the inflated terminology could complicate evidence-based reimbursement decisions by raising baseline expectations without raising baseline evidence.
Q: Is there a connection between the "Super Intelligence" rebrand and automation bias risk?
A: Yes, and it is direct. Automation bias forms faster when a tool is perceived as more capable than it actually is. "Super Intelligence" implies superhuman performance in a way that "artificial intelligence" does not. The rebrand, applied to tools whose performance data has not changed, may accelerate trust formation and reduce the scrutiny threshold — which is the mechanism by which automation bias develops. Physicians should treat the new label as clinically neutral while continuing to evaluate each tool on its actual evidence.
Dr. Adarsh K. Gupta, DO, MS is a clinical AI strategist and the author of The Science of Habits. He is a Cochrane Associate Editor with 25+ years of clinical experience. Follow him at adarshgupta.com and subscribe to the Clinical AI Brief for biweekly analysis of AI tools, regulations, and adoption decisions in clinical practice.