COVID, pressure and scientific dissent

David Zweig’s December 2022 thread described pressure from both the Trump and Biden administrations and examples of overbroad COVID enforcement. He highlighted Martin Kulldorff’s views on vaccination priorities and prior infection, and an internal review finding only one of Andrew Bostom’s five recorded violations valid. These are concrete examples to examine, not a reason to dismiss every complaint as partisan grievance. The thread also describes occasions when Twitter resisted pressure. [1]

His larger criticism was that moderators sometimes treated departure from public-health guidance as proof of falsity. Guidance combines scientific evidence with policy choices: who should be prioritized, how much risk is acceptable and how uncertainty should be communicated. A contested recommendation should be identified as such. Equally, possessing a medical credential does not automatically validate every statement a person makes. [1]

Actual vaccine harms should be stated plainly. CDC recognizes a rare causal association between mRNA vaccines and myocarditis/pericarditis, most often in adolescent and young adult males after a second dose. Reporting that risk is legitimate. But a verified rare risk does not establish that every illness after vaccination was caused by it. This is why a useful audit checks the specific words, evidence and date of each moderation decision instead of sorting speakers into permanently trustworthy or untrustworthy camps. [2]

Taibbi’s March 2023 reporting argued that the Virality Project helped police even true vaccine stories that might discourage vaccination. The criticism touched a real principle: people must be able to discuss documented harms without those harms being relabeled false merely because the discussion is inconvenient. The project was a research collaboration, however, not a government agency with direct power to delete tweets. [1][2]

A fuller exchange, reproduced by project participant Renée DiResta in screenshots and a June 2026 transcription, matters here. The March 17, 2021 email listed possible categories for Twitter to receive, including true material that might promote hesitancy; only the standard-misinformation category was marked as already being tagged. The next-day reply selected other categories rather than the true-content category. A proposed research category is not, on its own, proof of a demand to remove every item within it. [3]

That does not make the language beyond criticism. Treating true adverse-event narratives as a risk category can encourage a paternalistic approach to public debate. But a fair finding follows the full record: monitoring, escalating, labeling and deletion are different. DiResta’s rebuttal is a participant’s defense, not an independent verdict; the underlying email should be read alongside the original accusation rather than replaced by either side’s headline. [1][3]

Alex Berenson alleged that White House officials and Pfizer-connected figures pushed Twitter to suspend him over his vaccine reporting. The litigation contains correspondence relevant to that allegation, but a court recounting a complaint is not certifying that all alleged causation occurred. A July 2025 opinion dismissed claims against Andrew Slavitt, Scott Gottlieb and Albert Bourla on legal grounds including available relief, civil-rights pleading and failure to allege a contractual breach. [1]

On May 13, 2026, DOJ announced settlement of the government litigation. The new administration presented this as a free-speech victory and condemned Biden-era conduct. That is a consequential outcome. The announcement does not turn the government's political characterization into a trial finding, nor should a settlement be erased from the story because it complicates a dismissal narrative. [2]

On October 20, 2026, the Second Circuit affirmed dismissal of the remaining claims against Gottlieb and Bourla in a nonprecedential summary order. It held that the claimed unvaccinated class and alleged animus did not meet the invoked conspiracy statute’s requirements, and that no contract breach was adequately pleaded. The First Amendment challenge had become moot after stipulated dismissals involving the government defendants and others. These outcomes concern distinct defendants and legal theories; neither proves every complaint false or establishes that Pfizer ordered a ban. [3]