Abstract
COVID-19 created a rapidly changing evidence environment in which false claims could cause harm, but official positions could also become enforceable before their uncertainty was resolved. This paper distinguishes neutral lawfare—the strategic use of law—from suppressive lawfare, where process imposes disproportionate costs that deter speech or professional participation independently of the merits. It examines criminal communications enforcement, California AB 2098, AAPS v. ABIM, Apter v. HHS . The evidence does not support a single claim that all COVID dissent was suppressed or vindicated. It does support a disciplined investigation of how law, licensing, certification, employment and institutional amplification could convert provisional judgments into practical orthodoxy. The central empirical question is not simply who was right, but what was reasonably contestable at the time, what process was imposed and whether that process changed the behavior of others.
Executive summary
COVID-19 created a genuine governance dilemma. False medical claims could cause immediate harm, yet the evidence base changed quickly and official positions were sometimes revised. Institutions needed to act without converting provisional judgments into enforceable orthodoxy. The central question is therefore not whether misinformation existed—it plainly did—but whether the combined machinery for controlling it left adequate room for correction, dissent and good-faith disagreement.
An emergency information regime
In early 2020 public-health authorities had to communicate under radical uncertainty. Delay could cost lives, but premature certainty could also entrench error. Positions concerning transmission, population immunity, school closure, masking, vaccine performance and treatment evolved as evidence accumulated. Ordinary scientific correction—replication, criticism, conferences and gradual consensus—operated inside an emergency political environment demanding clear instructions.
The resulting information regime was decentralized but mutually reinforcing. Agencies issued guidance; professional bodies incorporated official positions; employers protected institutional reputations; journals corrected or withdrew papers; platforms reduced distribution; and critics sometimes faced civil or disciplinary proceedings. Coordination need not be proven for cumulative pressure to exist. Institutions can act independently from similar incentives and still produce a system in which the cost of dissent exceeds the cost of silence.
That does not make every intervention lawfare. A licensing board has a duty to investigate unsafe prescribing. An employer may enforce contractual obligations. A journal should correct unreliable work. A platform may address demonstrably dangerous falsehoods. The relevant questions are whether the rule was knowable, the authority was proper, the procedure was fair, the sanction was proportionate and the contested proposition was reasonably open at the time.
The distinction also corrects a common asymmetry. Litigation challenging a restriction and proceedings enforcing it are both strategic uses of law. Either may be legitimate; either may be abusive. The inquiry should therefore follow the dated claim, governing authority, procedure, evidence and effect rather than assign virtue according to which side invoked the court or regulator. COVID provides an unusually demanding test because rapidly changing knowledge made both premature enforcement and irresponsible contrarianism possible.
A taxonomy of pressure
| Mechanism | Legitimate purpose | Potential chilling pathway |
|---|---|---|
| Civil litigation | Protect legal rights, reputation and contractual interests. | Defense cost, discovery, injunction risk and compelled retraction. |
| Licensing discipline | Protect patients and professional standards. | Suspension, legal cost and loss of livelihood before final resolution. |
| Board certification | Signal competence and continuing standards. | Threat to credentials, hospital privileges and employability. |
| Employment action | Protect patients, staff and institutional mission. | Dismissal, affiliation loss and reputational signaling. |
| Journal process | Correct unreliable science and conflicts. | Investigation, retraction and reluctance to submit dissenting work. |
| Agency communication | Inform the public about evidence and authorized products. | Official denunciation may be treated as a binding professional rule. |
| Platform restriction | Limit fraud, dangerous falsehoods and manipulation. | Reduced reach, reputational labeling and hidden migration of debate. |
Lawfare need not be a pejorative term. Used descriptively, it means the strategic deployment of law or legal institutions to achieve wider political, economic or institutional objectives. Melillo’s history of international tobacco control illustrates that neutral usage: opposing networks used international law and evidence to advance or resist regulation. This paper is concerned with the narrower category of suppressive lawfare, in which legal or legal-administrative process imposes disproportionate collateral burdens that deter speech or participation regardless of the eventual merits. ‘Institutional pressure’ remains the broader category. Keeping the terms distinct prevents the paper from treating every adverse employment decision, publication dispute or platform intervention as legal abuse. (Dunlap, 2008; Melillo, 2024).
The communications-enforcement figures discussed in Paper 1 identify a broader proportionality question. They do not show how many cases concerned COVID-19, health policy or political disagreement, and the arrest, charge and conviction totals are not a matched cohort. This paper therefore does not use them as evidence of suppression of pandemic dissent (Big Brother Watch, 2026, pp. 48–57).
The temporal test
Retrospective certainty is the largest methodological trap. A claim later supported by evidence may have been reckless when made; an official position later revised may have been reasonable at the time. The proper unit is a dated proposition: the exact words used, the audience, the speaker’s professional role, the contemporaneous evidence, the institution’s stated standard and the sanction actually threatened or imposed.
Four clocks not one
Every case contains at least four timelines: the development of scientific evidence; the public position of the relevant agency; the speaker’s statements; and the initiation and resolution of the sanction or proceeding. These clocks often move at different speeds. An agency may retain a simplified message after its evidentiary basis has weakened, while a critic may state a plausible hypothesis with unjustified certainty. A fair evaluation must align all four before judging either the speech or the response.
The analysis should therefore avoid two mirror-image errors. Later vindication of part of a claim does not establish that every earlier formulation was responsible; later revision of official guidance does not prove that the original position was dishonest. Conversely, the existence of uncertainty does not authorize institutions to portray a contestable proposition as professional misconduct without a clear rule, fair procedure and proportionate response.
| Element | Question |
|---|---|
| Statement | What precisely was said, to whom, and was it policy advocacy, scientific interpretation or patient-specific advice? |
| Evidence at the time | What studies, uncertainties and authoritative reviews were available on that date? |
| Rule and authority | Which written standard applied, who had jurisdiction, and was the standard adopted before the conduct? |
| Process | Was notice specific, evidence disclosed, counsel permitted and independent appeal available? |
| Proportionality | Was the response connected to demonstrated harm and narrower alternatives? |
| Effect | Did the speaker or peers alter practice, research, publication or public participation? |
California AB 2098
California’s AB 2098 designated physician dissemination of COVID-19 misinformation or disinformation to patients as unprofessional conduct. The statute linked professional discipline to a definition referencing contemporary scientific consensus and the standard of care. Physicians challenged it on constitutional grounds, arguing that the definition was vague and chilled doctor-patient discussion. The law was later repealed, and the Ninth Circuit dismissed the principal appeal as moot. (California Legislature, 2022).
The sequence supports a limited conclusion. A state attempted to make disputed COVID-19 claims professionally enforceable during a rapidly changing evidence environment, and repeal left the principal appeal moot rather than producing a final appellate decision on the constitutional merits. It does not follow that every statement by every plaintiff was protected or correct. Nor does repeal establish that enforcement actually occurred. The strongest empirical inquiry is whether physicians changed consultations, consent materials or public speech while the law was in force, and whether those changes are documented contemporaneously. (McDonald v. Lawson, 2024).
AB 2098 also shows law operating in both directions. The statute created disciplinary exposure; constitutional litigation challenged that exposure. Both sides described the other as endangering patients or speech. The appropriate test concerns institutional design rather than choose a side by label: Was the prohibited proposition sufficiently specific? Could ordinary malpractice rules address patient harm? Did the rule distinguish public-policy speech from treatment of an identifiable patient?
Certification and threatened loss AAPS v ABIM
In AAPS v. ABIM, an educational foundation alleged that medical specialty boards threatened decertification of physicians expressing contested views about COVID-19 and other issues. The district court dismissed the case. In June 2024 the Fifth Circuit held that AAPS had sufficiently alleged standing and should have been allowed to amend claims against the board defendants; it affirmed mootness as to the dissolved Department of Homeland Security Disinformation Governance Board. (AAPS v. ABIM, 2024).
The appellate ruling is important but narrow. It accepted at the pleading stage that threatened decertification could predictably induce self-censorship and injure an organization whose members withheld speech. It did not determine that the boards violated the Constitution or antitrust law, that particular physicians were entitled to certification, or that their medical claims were correct. The distinction belongs prominently in any publication because standing decisions are often misreported as merits vindication. (AAPS v. ABIM, 2024).
Certification nonetheless presents an unusually powerful private lever. It is formally voluntary, yet can affect hospital privileges, insurance participation and employment. Research should identify the published misinformation policies, notices sent, final sanctions, appeal routes and evidence linking disputed speech to credential loss. It should also interview physicians who received no notice but altered speech after observing action against others.
Agency speech and statutory limits Apter v HHS
Apter arose from FDA communications discouraging ivermectin use for COVID-19, including highly compressed social-media messages. Physicians alleged that the campaign interfered with their practice and contributed to professional consequences. The Fifth Circuit reversed a jurisdictional dismissal in 2023, holding that the physicians could pursue an ultra vires theory and had plausibly alleged that some posts crossed from informing the public into recommending how a medicine should be used. (Apter v. HHS, 2023).
Again, the holding was institutional rather than clinical. The court did not decide that ivermectin was effective against COVID-19. It examined statutory authority and sovereign immunity. That separation is analytically valuable: an agency may communicate a defensible scientific conclusion in a manner that exceeds its lawful role, while a critic may win a jurisdictional point without proving the treatment claim. (Apter v. HHS, 2023).
The case also demonstrates amplification. An informal post can be repeated by hospitals, pharmacies, boards and journalists until it functions like a rule, even though it creates no direct legal obligation. A defensible study would trace citations: which later decisions expressly relied on FDA wording, which acted independently, and whether fuller agency language produced different downstream effects than social-media shorthand. The case subsequently ended through a 2024 settlement addressing specified communications. That resolution did not establish clinical efficacy.
Individual cases must remain individual
Claims about named clinicians must be reconstructed individually from primary records. A dispute over prescribing, patient records or a former employer’s affiliation rules cannot be treated as a case about public-policy speech without evidence connecting the two. This paper confines its legal comparisons to the identified decisions and does not adjudicate unresolved allegations against individual doctors.
Patient-specific misconduct can coexist with legitimate policy criticism, and institutional overreach can coexist with weak scientific claims. Evaluating the statement, professional duty and sanction separately gives any finding of retaliation a firmer foundation.
| Matter | Documented procedural result | Limit of the inference |
|---|---|---|
| California AB 2098 and McDonald v Lawson | Repeal rendered the principal appeal moot in February 2024 | No final appellate merits decision on the constitutional claims |
| AAPS v ABIM June 2024 | Standing and opportunity to amend addressed at the pleading stage | No finding that every alleged sanction was unlawful or every medical claim correct |
| Apter v HHS September 2023 | Ultra vires theory survived sovereign-immunity dismissal; later settled | No judicial finding that ivermectin treats COVID-19 effectively |
Measuring cumulative burden
The hypothesized harm is not only the sanction imposed on a named physician. It is the effect on the surrounding professional population. Useful indicators include declined manuscripts, canceled lectures, withdrawn grant proposals, changes to consent discussions, reluctance to study disfavored hypotheses, private emails contradicting public statements, movement to less visible platforms and colleagues’ contemporaneous accounts of self-censorship.
Duration should be measured separately from outcome. A suspension later reversed can still eliminate income and patients. An investigation ending without discipline can still require counsel and years of uncertainty. A successful plaintiff may recover fees but not reputation or lost research time. Conversely, a highly publicized proceeding may amplify the challenged speech. The net effect is an empirical question.
Comparison is essential. Researchers should identify physicians exposed to similar institutional policies who continued speaking, and clinicians disciplined for clearly patient-specific conduct unrelated to public dissent. Without such controls, every career setback risks being attributed to censorship and every absence of speech to fear.
From an individual proceeding to a chilling effect
A chilling effect is a population claim and should not be inferred from the burden on one prominent individual. Evidence should include contemporaneous accounts from people who changed conduct: a physician who removed language from consent materials, a researcher who abandoned a question, an editor who declined a submission, or an employee who stopped speaking after observing a colleague’s treatment. Whenever possible, those accounts should be compared with similarly placed professionals who continued to speak.
Useful measures include the number and duration of investigations, interim suspensions, defense expenditure, lost clinical income, withdrawn manuscripts, canceled lectures, altered institutional policies and changes in public participation before and after a proceeding. Anonymous interviews can identify the mechanism; publication should distinguish them from documentary proof and from named, corroborated testimony.
Institutional comparisons
The limits of borrowed authority
The same analytical problem can arise outside COVID-19. Broad scientific findings do not automatically settle narrower claims about a particular event, exposure or individual. Paper 2 examines this issue in product toxicology and climate attribution. Here the relevant test is whether a professional institution supplied the additional reasoning needed to move from a general finding to a sanction against a particular speaker.
Private funding inside public enforcement offices
The preceding cases concern professional participation during COVID. A comparative example from environmental enforcement raises a related question: who finances the capacity through which public authority is exercised? NYU announced its State Energy and Environmental Impact Center in 2017 with initial Bloomberg Philanthropies support and a fellowship program placing lawyers in state attorneys general’s offices. The Center’s account says NYU paid fellows’ salaries, while receiving officials directed their work and held their sole loyalty. Those safeguards matter; private payment alone does not establish improper donor direction. (NYU School of Law, 2017; State Energy and Environmental Impact Center, 2026a).
A July 2025 House Oversight Committee investigation questioned the NYU fellowship arrangement. That was an investigation, not an adjudication of wrongdoing. The appropriate inquiry concerns actual grant terms, public supervision, appointment decisions and donor influence over individual proceedings. An archived website or a critical essay cannot establish why a programme ended or whether misconduct occurred (US House Committee on Oversight and Government Reform, 2025).
The testable concern is selection rather than presumed misconduct. Funding restricted to a policy field can expand enforcement capacity there while leaving competing priorities dependent on ordinary public budgets. Evaluation should examine grant conditions, appointment agreements, case selection, supervision, renewal incentives and donor communications. It should then apply the same proportionality test used elsewhere in this paper: what claim justified the proceeding, what burden followed and what remedy was sought? Philanthropic support does not itself prove suppressive lawfare, but charitable status cannot exempt the machinery of enforcement from scrutiny.
Research programme
| Workstream | Evidence required | Output |
|---|---|---|
| Case chronologies | Pleadings, board notices, contracts, policies, decisions, appeals and dated statements. | Verified narratives separating allegation, procedural ruling and merits finding. |
| Evidence timelines | Contemporaneous studies, official guidance, revisions and uncertainties. | Assessment of whether a proposition was false, contestable or well supported when made. |
| Chilling survey | Confidential interviews plus documents demonstrating behavioral change. | Anonymized aggregate findings with named cases only where corroborated and authorized. |
| Institutional network | Citations among agencies, boards, employers, journals and platforms. | Map of independent action versus documented reliance or coordination. |
| Counterexamples | Cases protecting patients, correcting fraud or leaving dissent untouched. | Tests against selection bias and overbroad claims. |
The research programme above is prospective. Its proposed surveys, interviews and case dossiers are not reported findings. Publication of future results should identify selection rules, evidence dates, counterexamples and the distinction between recollection and contemporary records.
Safeguards
Emergency speech-related rules should expire unless renewed after evidence and public review. Licensing and certification standards should distinguish patient-specific care from public-policy speech, identify the exact proposition considered false, state the evidence date and provide independent appeal. Employers should separate contractual affiliation claims from scientific disagreement. Agencies should preserve context when communicating through social media and disclose the evidentiary basis and limits of their advice.
Where private funding supports personnel exercising public authority, the funding terms, supervision arrangements and relevant conflicts should be disclosed. Governments should control appointments and case selection, prohibit donor direction of individual proceedings, and preserve records and avenues of review after a grant or sponsored project ends. These standards should apply to charitable and commercial funders alike.
These safeguards do not create a right to mislead patients or a guarantee of employment, publication or amplification. They make sanctions reviewable and proportionate. The objective is a system capable of addressing dangerous falsehoods without making institutional conformity the substitute for scientific reasoning.
Conclusion
The COVID period does not sustain the simple claim that all dissent was suppressed, that all institutional action was coordinated or that every sanctioned doctor was vindicated. It does sustain a serious inquiry into cumulative professional coercion. The most persuasive cases will show that a dated, reasonably contestable proposition triggered burdens disproportionate to patient harm and that those burdens changed the behavior of people beyond the immediate target.
The method separates scientific accuracy from institutional authority, procedural victories from merits decisions, and individual burden from population effects. That discipline is necessary to identify coercion rather than merely express disagreement with institutions. A finding of suppressive lawfare requires evidence of disproportionate process and its effects.
Relationship to the series
Paper 3 applies Paper 1's process test to individuals and professional institutions rather than mass corporate defendants. Paper 2 explains how a proposition first acquires scientific authority and, through the climate-attribution comparison, how litigation demand may also shape the production of evidence upstream. Paper 4 then asks how comparable pressures operate in international health governance where access, funding and reputation may substitute for a conventional lawsuit. Paper 5 considers the organizational incentives that may sustain these pressures.
NYU School of Law (2017) NYU Law launches new center to support state attorneys general in environmental litigation. Available at: https://www.law.nyu.edu/news/state-impact-center-attorneys-general-environmental-litigation-climate-energy (accessed September 30, 2026).
State Energy and Environmental Impact Center (2026a) Fellows Program. Available at: https://stateimpactcenter.org/about/fellows-program (accessed September 30, 2026).
US House Committee on Oversight and Government Reform (2025) Comer investigates Bloomberg Philanthropies for partisan activism with state attorneys general, July 17. Available at: https://oversight.house.gov/release/comer-investigates-bloomberg-philanthropies-for-partisan-activism-with-state-attorneys-general/ (accessed September 30, 2026).
AAPS v. ABIM, No. 23-40423 (5th Cir. 3 June 2024). Available at: https://www.ca5.uscourts.gov/opinions/pub/23/23-40423-CV0.pdf (Accessed: 7 September 2026).
Apter v. Department of Health and Human Services, 80 F.4th 579 (5th Cir. 2023). Available at: https://www.ca5.uscourts.gov/opinions/pub/22/22-40802-CV0.pdf (Accessed: 7 September 2026).
