Gain-of-Function, Defined and Redefined

I. Three words, two meanings

On May 11, 2021, in a hearing room of the United States Senate, two men used the same three words to mean two different things.

Senator Rand Paul said the Wuhan Institute of Virology had been “experimenting to enhance the coronavirus’ ability to infect humans,” that this was “gain-of-function research,” and that the NIH had funded it.¹ Anthony Fauci answered, with finality: “it is not gain-of-function. And, if you look at the grant and you look at the progress reports, it is not gain-of-function.”¹

To most people watching, that was a flat denial that the United States had paid to make a virus more dangerous. Three years later, under oath and without a clock running, Fauci explained that his denial had never meant that. It meant something far more specific — and the difference is the whole story.

II. Three definitions

In his January 2024 transcribed interview, congressional staff did something the cable-news exchanges never had time for: they slowed down and asked Fauci to define his terms. He agreed that “gain-of-function” had been used in at least three distinct ways.²

The three are not interchangeable. The first sweeps in almost any genetic modification that adds a capability. The third is a narrow regulatory trigger: it applies only to a potential pandemic pathogen — one already capable of seeding a pandemic — and only when an experiment is “reasonably anticipated” to make it more transmissible or more lethal. A great deal of work is “gain-of-function” under the first definition and not under the third.

III. The operative definition

Fauci was explicit about which one he uses. The P3CO framework, he testified, came out of years of deliberation and “a very precise definition”: an experiment “reasonably anticipated to result in the enhancement of” a potential pandemic pathogen.³

And then the pivot the public never heard:

“when I said we do not fund gain-of-function — gain-of-function according to the strict definition, which I refer to as the operative definition… when someone asks me, as a scientist, are you doing gain-of-function, is that gain-of-function, I always apply it to the operative definition of ‘gain-of-function.’”⁴

This deserves to be read charitably, because it is defensible. Fauci is a scientist and a regulator; when a regulatory term has a precise legal meaning, applying that meaning is not evasion — it is precision. The P3CO definition is real, it was the binding federal standard, and it is the definition that determined whether a given experiment required extra review. A researcher answering “are you doing gain-of-function?” with the regulatory definition in mind is doing something legitimate.

The difficulty is not that Fauci used the narrow definition. It is that, when he used it in public, he did not say he was using it.

IV. What the public heard

Return to the Senate hearing with the definitions in hand, and the exchange reads very differently.

Paul was plainly invoking the broad meaning — that the WIV had worked to “enhance the coronavirus’ ability to infect humans” and “create super viruses.”¹ Fauci answered with the narrow one: “it is not gain-of-function.”¹ Both statements can be true at once. Under Paul’s definition, inserting spike genes to give a virus new ability to infect human cells is, self-evidently, a gain of function. Under Fauci’s operative definition, the question is whether the result was a potential pandemic pathogen enhanced to pose pandemic risk — a far higher bar the NIH says was not met.

But a viewer could not have known that two definitions were passing in the night. They heard a categorical denial, in plain English, that the government had funded the enhancement of a dangerous virus.

V. The redefinition

The title of this piece is not rhetorical. The operative definition really was redefined — by the government, on the record.

The 2014 moratorium was a blunt instrument: a funding pause on work that might give influenza, MERS, or SARS viruses “enhanced pathogenicity and/or transmissibility… via the respiratory route.”⁵ In 2017 it was replaced by the P3CO framework, which Fauci described as introducing “an even more specific set of definitions: potential pandemic pathogen, which is a multipart definition, and all sorts of carve-outs.”² The standard narrowed: from a broad class of dangerous viral enhancements to a tightly bounded category of pandemic-capable pathogens.

That is a legitimate policy evolution — narrowing a rule so it captures the genuinely catastrophic cases is a defensible regulatory choice. But it had a consequence for public understanding. As the definition narrowed, the set of experiments an official could truthfully say were “not gain-of-function” grew — even as the everyday meaning of the phrase, the one the public carried, did not budge.

VI. Does the work fit?

So: was the Wuhan work “gain-of-function”? The only honest answer is under which definition.

Under the broad one, the grant’s own progress report answers the question. As this series has documented, NIH-funded researchers built chimeric coronaviruses and found that one made humanized mice markedly sicker — “the pathogenicity of SHC014 is higher,” in the report’s words.⁶ Conferring enhanced pathogenicity is a gain of function under the layman’s definition by construction.

Under the operative definition, the NIH says no: in its October 2021 letter to Congress, it stated the work “was not subject to departmental review under the HHS P3CO Framework,” because the viruses studied were not potential pandemic pathogens and “could not have become SARS-CoV-2.”⁷ Both of those statements can be true. They are answers to different questions.

VII. What the record establishes — and what it does not

The record establishes:

• That “gain-of-function” carried at least three distinct meanings, by Fauci’s own testimony, and that he “always” answered with the narrowest, regulatory one.²⁴
• That in public, in 2021, he denied funding “gain-of-function” research without indicating that his denial rested on that specific definition.¹
• That the operative definition had itself been narrowed in 2017, shrinking what could be denied.²⁵

The record does not establish — and we do not assert — that:

• Fauci lied. Applying a precise regulatory definition is legitimate; the masthead’s name is not this article’s verdict.
• The narrow definition was invented to dodge the question. The P3CO framework predates the controversy by years and was the binding federal standard.
• The broad critics are simply correct, or that the work met the threshold for pandemic risk. The NIH’s contrary position is on the record and represented here.

The failure the documents show is narrower, and it is about candor: a public yes-or-no answer that secretly turned on a definition the audience was never given.

VIII. Why it matters

A term of art is a fine thing in a regulatory filing. It is a dangerous thing at a podium. When an official answers a plain public question — did we pay to make this virus more dangerous? — with a word that he is silently defining more narrowly than his audience, the public is not lied to so much as quietly disqualified from the conversation. They are given a true sentence they cannot evaluate.

The remedy was never a verdict on virology. It was one clause: by the regulatory definition. Said aloud, in 2021, it would have turned a categorical denial into an honest, contestable claim. Left unsaid, it turned a definition into a place to hide.