Gain-of-Function Benefits Outweigh Risks

• Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases for nearly four decades, was the most consequential institutional champion of the view that gain-of-function research was worth its risks. In a 2012 paper published in mBio, he argued plainly that the benefits of such experiments outweighed the dangers, reasoning that a natural pandemic was far more likely than a laboratory accident and that scientists needed to stay ahead of the virus. ^[2] That argument became the intellectual foundation for NIAID's funding decisions for years afterward, including grants channeled through EcoHealth Alliance to the Wuhan Institute of Virology for bat coronavirus research. ^[1] When COVID-19 arrived and questions about its origins began circulating, Fauci remained the public face of the American pandemic response, a position that made scrutiny of his earlier funding decisions politically charged and institutionally awkward. ^[5]
• Peter Daszak, president of EcoHealth Alliance, occupied a peculiar position in the gain-of-function debate: he was simultaneously a leading advocate for the research, a primary conduit for federal funds flowing to the Wuhan Institute of Virology, and one of the most vocal dismissers of the laboratory origin hypothesis for SARS-CoV-2. ^[1] In 2018, his organization submitted a proposal to DARPA, code-named DEFUSE, that outlined plans to insert furin cleavage sites into bat coronaviruses at a BSL-2 facility, a biosafety level roughly equivalent to a dentist's office. ^[6] When the pandemic began, Daszak organized and signed a letter in The Lancet condemning lab-leak theories as conspiracy thinking, without disclosing his organization's direct involvement in coronavirus research at the Wuhan lab. ^[1] DARPA had already rejected the DEFUSE proposal on biosafety grounds, a fact that did not become public until the document leaked years later. ^[6]
• Ron Fouchier of Erasmus University and Yoshihiro Kawaoka of the University of Wisconsin were the researchers who, in 2011 and 2012, independently demonstrated that H5N1 influenza could be made airborne between ferrets through a small number of mutations. ^[5] The experiments were celebrated by proponents as proof that gain-of-function research could identify pandemic threats before nature did. Critics saw the same results differently: the work had just demonstrated, in a laboratory, how to make one of the world's most lethal bird flu strains transmissible between mammals. The publications triggered a voluntary moratorium and a prolonged, unresolved debate about whether the knowledge gained justified the risk created. ^[2]
• Marc Lipsitch, professor of epidemiology at Harvard T.H. Chan School of Public Health, was among the most persistent and technically specific critics of the assumption. He argued that laboratory-induced mutations do not reliably predict natural viral evolution, that results from one strain rarely generalize to others, and that the claimed benefits for outbreak preparedness were largely speculative. ^[3] His position was that there was no demonstrated route by which gain-of-function studies had answered, or could answer, the key questions that arise during an actual outbreak, and that the research community was confusing the production of interesting findings with the production of useful ones. ^[3] Richard Ebright, a molecular biologist at Rutgers University, made a complementary argument from the regulatory side, criticizing the NIH's definitions of gain-of-function as so narrow and malleable that they excluded research that plainly qualified, including the chimeric coronavirus experiments conducted at Wuhan. ^[5] When the DEFUSE proposal leaked, Ebright noted that the planned insertion of furin cleavage sites into bat coronaviruses was directly relevant to the distinctive feature of SARS-CoV-2 that had puzzled virologists since the pandemic began. ^[6]
• Alina Chan, a scientist at the Broad Institute, became one of the more prominent voices arguing that the DEFUSE proposal materially shifted the probability calculus toward a laboratory origin for SARS-CoV-2. ^[6] Her position was not that the lab origin was proven, but that the proposal's existence made it logically consistent in a way that the scientific establishment had been too quick to foreclose. Christophe Fraser, Imperial College's deputy director for outbreak analysis, raised a narrower but pointed objection at a National Academies symposium: even granting the research its claimed benefits, he said the value of gain-of-function data during actual outbreaks was unclear, and that prediction failures had resulted from surveillance gaps rather than any shortage of laboratory-derived transmissibility data. ^[4]

The National Institute of Allergy and Infectious Diseases was the primary federal engine behind gain-of-function research funding in the United States. Under Fauci, NIAID supported H5N1 transmissibility studies, endorsed a temporary moratorium in 2012 for policy review while continuing to back related work, and channeled grants through EcoHealth Alliance to the Wuhan Institute of Virology during the 2014 funding pause that was nominally intended to halt exactly this category of research. ^[2][5] The agency's position throughout was that the science was sound, the oversight was adequate, and the benefits to pandemic preparedness were real. Congressional scrutiny and FOIA disclosures later raised serious questions about whether the pause had been honored in practice. ^[5]

EcoHealth Alliance functioned as the institutional intermediary that connected American federal funding to high-risk coronavirus research in China. The organization channeled NIAID grants to the Wuhan Institute of Virology for experiments involving chimeric bat coronaviruses, and in 2018 submitted the DEFUSE proposal to DARPA outlining plans to engineer furin cleavage sites into bat coronavirus backbones at a BSL-2 laboratory. ^[6] DARPA rejected the proposal, citing gain-of-function concerns, but the work it described was consistent with research that had already been funded through other channels. ^[6] The organization's dual role, as both a recipient of federal biosecurity funding and a vocal participant in the debate over COVID-19's origins, became a central point of controversy after the pandemic began. ^[1]

The CDC incorporated gain-of-function-derived data into its Influenza Risk Assessment Tool, the instrument used to prioritize which emerging strains warranted the most urgent attention and resource allocation. ^[4] This institutional embedding of gain-of-function outputs into operational public health decision-making was cited by proponents as concrete evidence of the research's practical value. It also meant that the credibility of those risk assessments was, at least in part, tied to the credibility of the underlying experimental methodology. St. Jude Children's Research Hospital, operating as both an NIAID-funded Center of Excellence and a WHO collaborating center for influenza surveillance, used its position in the global monitoring network to argue that gain-of-function experiments were essential for identifying zoonotic threats and selecting vaccine strains. ^[4] The same CDC that housed this tool also, in 2014, accidentally contaminated benign influenza samples with H5N1 and exposed 75 staff members to anthrax through procedural failures, incidents that did not visibly alter the institutional consensus on the manageability of laboratory risk. ^[5][7]

The core argument for gain-of-function research rested on a straightforward claim: that scientists needed to know what a pandemic pathogen might look like before nature produced it. If a few targeted mutations could make H5N1 airborne between ferrets, the reasoning went, then identifying those mutations in the laboratory gave public health authorities a head start on vaccines and countermeasures. Fauci made this case explicitly in his 2012 mBio paper, arguing that the risk of a laboratory accident was lower than the risk of a natural pandemic and that the knowledge produced by such experiments was essential to preparedness. ^[2] The argument had genuine force in the context of H5N1, a virus with a case fatality rate above 50 percent in humans that had not yet acquired efficient human-to-human transmission. The question of whether it ever would, and what that would look like, was not academic.

The H5N1 ferret experiments by Fouchier and Kawaoka appeared to validate the framework. They showed that a small number of mutations, some of which had already been observed in nature, were sufficient to confer airborne transmissibility in mammals. ^[5] Proponents argued this was exactly the kind of advance warning the research was designed to provide. At a National Academies symposium, researchers cited additional short-term benefits: gain-of-function techniques adapted viruses to laboratory culture, generated animal models for emerging pathogens like MERS-CoV, and illuminated escape mutations relevant to vaccine design. ^[4] Schultz-Cherry of St. Jude argued that the research revealed transmissibility markers that could guide vaccine prioritization, pointing to the Cambodia H5 outbreak as a case where such data had informed CDC risk tools. ^[4]

The EcoHealth Alliance's DEFUSE proposal rested on a narrower but related claim: that inserting novel furin cleavage sites into bat coronaviruses would allow researchers to evaluate the growth potential of chimeric viruses and assess spillover risk before natural evolution produced the same result. ^[6] Proponents of the work, including Vincent Racaniello of Columbia University, argued that the benefits of such research far outweighed the risks. ^[6] The proposal classified the work as exempt from gain-of-function concerns on the grounds that it used bat coronaviruses rather than known human pathogens, a distinction that critics found unpersuasive given the demonstrated ability of such viruses to infect human cells. ^[6] Growing evidence now suggests this distinction was less meaningful than its proponents claimed, though the debate over what exactly constitutes gain-of-function research, and who gets to decide, has never been fully resolved. ^[5]

The assumption also drew credibility from a genuine epistemic problem: researchers cannot reliably predict a virus's phenotype from its genotype alone. Schultz-Cherry cited this limitation as a reason why ongoing gain-of-function experiments were necessary, arguing that until the genotype-phenotype relationship was better understood, empirical testing was the only available tool. ^[4] Critics like Lipsitch turned the same observation around: if evolution is sufficiently random that laboratory mutations do not predict natural ones, then the research's claimed predictive value is undermined by the same uncertainty it was designed to resolve. ^[3] The Wuhan chimeric coronavirus experiments, in which bat spike proteins on a WIV1 backbone infected human cells and produced viruses ten thousand times more concentrated in mouse lungs than the parent strain, illustrated the problem concretely: the experiments generated striking results, but whether those results were informative about natural pandemic risk or simply demonstrated that dangerous chimeras could be created remained genuinely contested. ^[5]

The assumption spread most effectively through the funding and publication structures of academic virology. NIAID grants shaped research agendas, and researchers whose work depended on those grants had institutional reasons to accept the framework that justified them. The National Academies convened a workshop that amplified the benefits case through presentations on surveillance, vaccine development, and zoonotic threat assessment, treating gain-of-function as a category of research whose value was broadly comparable to undisputed biomedical advances. ^[4] The workshop format, with its mix of proponents and critics, gave the proceedings an appearance of balanced deliberation while the institutional weight of the convening body lent authority to the conclusions.

When COVID-19 arrived and the laboratory origin hypothesis began gaining traction, the scientific journals that had published the original gain-of-function research became a front in the debate over origins. A letter organized by Daszak and published in The Lancet in February 2020 condemned lab-leak theories as conspiracy theories and misinformation, framing the question as settled before any serious investigation had occurred. ^[1] Papers in Cell and other leading journals argued there was no evidence of prior furin cleavage site insertion research, a claim that the later leak of the DEFUSE proposal directly contradicted. ^[6] The effect was to make dissent from the natural-origin consensus professionally costly at precisely the moment when the question most needed open examination.

A persistent definitional dispute compounded the confusion. The NIH's working definition of gain-of-function was narrow enough to exclude much of the research that critics considered most dangerous, including the chimeric coronavirus work at Wuhan. ^[5] This created a situation in which the agency could truthfully say it had not funded gain-of-function research as it defined the term, while critics argued the definition had been drawn specifically to exclude the work in question. The resulting rhetorical standoff, amplified by congressional hearings and media coverage, produced more heat than resolution and left the underlying risk-benefit question largely unaddressed in public discourse. ^[5]

The first major policy response to the gain-of-function debate was a voluntary moratorium declared by leading H5N1 researchers in January 2012, following the controversy over the ferret transmission papers. The pause was intended to allow time for a transparent public discussion of the risks and benefits before the research resumed. ^[2] NIH and the U.S. government simultaneously moved to expand the dual-use research of concern oversight framework to cover fifteen pathogens, an acknowledgment that the existing review structures were inadequate for the category of work being done. ^[2] The moratorium lasted roughly a year and ended without any fundamental change to the regulatory framework.

In October 2014, the Obama administration imposed a broader funding pause on gain-of-function research involving influenza, MERS, and SARS that could enhance transmissibility or virulence in mammals, affecting eighteen laboratories. ^[5][7] The pause followed a series of biosafety incidents at federal facilities, including the CDC anthrax exposure and the H5N1 contamination of benign flu samples, that had made the argument for adequate containment harder to sustain. ^[5] NIH allocated millions to EcoHealth Alliance for bat coronavirus research during this period, with the agency later maintaining that the work did not meet the definition of gain-of-function research subject to the pause, a position that congressional investigators disputed. ^[1][5]

The HHS established a formal review process in 2017 for research that could enhance pathogen transmissibility or virulence, and the White House issued a unified policy in 2024 that narrowed the oversight framework to focus on the highest-risk pathogens. ^[7] At the state level, Florida banned gain-of-function research outright in 2023, and the Trump administration began considering a broader federal funding pause for research involving enhancements to deadly or highly contagious pathogens. ^[7] DARPA's 2018 rejection of the DEFUSE proposal on biosafety grounds represented an earlier instance of a federal agency applying exactly the kind of precautionary judgment that critics argued the NIH had consistently failed to exercise. ^[6]

The assumption began to unravel through a series of disclosures that made the official risk-benefit calculus harder to defend. DARPA's rejection of the DEFUSE proposal in 2018, which did not become public until the document leaked in 2021, revealed that at least one federal agency had looked at EcoHealth Alliance's planned coronavirus engineering and concluded the gain-of-function risks were unacceptable. ^[1][6] The proposal's description of inserting furin cleavage sites into bat coronaviruses at a BSL-2 facility, combined with SARS-CoV-2's distinctive furin cleavage site, made the laboratory origin hypothesis logically coherent in a way that the scientific establishment's early dismissals had not accounted for. ^[6] Ebright and Chan were among the first credentialed scientists to state this publicly, at professional cost. ^[6]

Congressional hearings and FOIA disclosures through 2021 and 2022 exposed the definitional disputes that had allowed NIAID to fund chimeric coronavirus research at Wuhan while maintaining it had not funded gain-of-function work subject to the 2014 pause. ^[5] The exchanges between Fauci and Senator Rand Paul made the definitional question visible to a general audience for the first time, though they produced more accusation than resolution. The NIH eventually acknowledged in a letter to Congress that EcoHealth Alliance had failed to report results showing enhanced pathogen growth, a limited concession that critics argued understated the significance of what had been funded. ^[5]

At the scientific level, an influential minority of researchers increasingly argued that the foundational claims for gain-of-function research had not held up. Lipsitch and colleagues contended that the evolutionary randomness of viral mutation meant laboratory results did not generalize reliably, that surveillance gaps rather than missing gain-of-function data had caused pandemic prediction failures, and that alternative approaches offered better returns at lower risk. ^[3] The National Academies symposium had acknowledged, without resolving, that the inability to predict phenotype from genotype undermined the research's core promise, and that vaccine manufacturers themselves disagreed about whether gain-of-function was essential to their work. ^[4] The debate is not fully settled, and proponents continue to argue that the research has produced genuine advances. But the confident institutional consensus that once made the risk-benefit calculation seem obvious has not survived contact with the evidence that accumulated after 2019. ^[7]

• Affirmative Action Causes No Reverse DiscriminationBusiness Civil Rights Economy Public Health Science Policy
• Airport Profiling is Racial DiscriminationBusiness Civil Rights Economy Public Health
• Anti-Bias Training WorksBusiness Civil Rights Economy Public Health
• DEI Drives Business SuccessBusiness Civil Rights Economy Public Health
• Policing Disparities Prove DiscriminationBusiness Civil Rights Economy Public Health