For several days in December 2021, Twitter's public COVID-19 misinformation policy stated effectively that it would punish users who claimed that vaccinated people could "spread or shed the virus" to others, a claim that is in fact accurate. However...
Twitter says its actual policy always related to claims that vaccinated people could spread the vaccine to others. According to the company, the use of the word "virus" in that context was merely a typo, which was corrected on Dec. 16, and did not reflect the true policy anyway.
In December 2021, a controversy emerged over the wording of one of Twitter's policies against COVID-19 related misinformation, with critics of the platform, many of them also vaccine opponents, claiming that it penalized users who posted even accurate statements.
On Dec. 11, for example, the website Reclaim the Net published a blog post with the headline "Twitter will now ban users that repeatedly claim vaccinated people can spread Covid." That post read:
Twitter has quietly updated its “COVID-19 misleading information policy” to impose new sanctions on tweets about vaccines, PCR tests, and health authorities. These sanctions include removing and labeling tweets. Both types of sanctions also result in Twitter users accruing strikes on their account which can lead to a permanent suspension.
...One of the most notable changes to this “COVID-19 misleading information policy” we noticed is related to claims about whether vaccinated people can spread the coronavirus. The policy now states that Twitter will label tweets with “corrective information” and give users a strike if they:
- Claim that “the vaccines will cause you to be sick, spread the virus, or would be more harmful than getting COVID-19”
- Post what Twitter describes as “false or misleading claims that people who have received the vaccine can spread or shed the virus (or symptoms, or immunity) to unvaccinated people”
This means Twitter users could now be sanctioned for sharing or discussing the Centers for Disease Control and Prevention’s (CDC’s) admission that “vaccinated people can still become infected and have the potential to spread the virus to others.”
Several right-leaning websites subsequently posted similar reports, each of them accurately pointing out that vaccinated individuals can, in fact, spread COVID-19 to others, as the U.S. Centers for Disease Control and Prevention (CDC) has indeed stated (although CDC guidance stipulates, importantly, that vaccinated people spread the virus at much lower rates than unvaccinated people do).
Those articles accurately represented the text of Twitter's COVID-19 misinformation policy, as it appeared for several days in December 2021. However, the company changed the wording on Dec. 15, and a spokesperson told Snopes the previous version was no more than a typo. More importantly, the spokesperson said the company had never, in its actions, treated as false information the claim that vaccinated people can spread COVID-19. As a result, we are issuing a rating of "Mixture."
What Twitter's policies say about COVID-19 transmission and vaccination
Over the course of the COVID-19 pandemic, Twitter has revised and refined its policies against potentially harmful misinformation around the virus, its spread, and the range of vaccines made available since December 2020. Based on various archived versions of that "COVID-19 misleading information policy," the following is a breakdown of the "vaccinated transmission" policy in question.
When Tweets include misleading information about COVID-19, we may place a label on those Tweets that includes corrective information about that claim. In cases where we determine there is potential for harm associated with the misleading claim, we will disable the ability for others to Retweet, Quote Tweet, or engage in other ways to prevent the spread of the misleading information.
...We may apply labels to Tweets that contain, for example:
...3. False or misleading information regarding the safety or science behind approved or authorized COVID-19 vaccines, such as:...
- False or misleading claims that people who have received the vaccine can spread or shed the virus (or symptoms, or immunity) to unvaccinated people. [Emphasis is added]
So false information would be labeled as such, with one kind of false information being that related to the safety of approved vaccines, with the specific example of the claim that vaccinated people can spread the virus to unvaccinated people — a claim that is, in fact, true.
And while Twitter would not ban a user outright for violating this policy just once, under the company's "strikes" system, repeatedly violating the false information policy in this way would ultimately lead to a permanent suspension.
That wording remained in place until the evening of Dec. 15, when Twitter changed it again, replacing the word "virus" with "vaccine," so that the example of false information now read as follows:
- False or misleading claims that people who have received the vaccine can spread or shed the vaccine (or symptoms, or immunity) to unvaccinated people. [Emphasis is added]
The claim that vaccinated individuals can "shed" the vaccine is in fact false, as Snopes has outlined in greater detail before.
As a simple visualization of the change that was made on Dec. 15, the screenshot below shows the text of this part of the policy, as it appeared on Dec. 10:
On Dec. 16, a Twitter spokesperson told Snopes, in an email, that the earlier "virus" wording was simply a typo, and didn't reflect the policy that was actually enforced by the company. The spokesperson wrote:
The earlier language is not reflective of our enforcement approach, past or present. We’ve enforced our COVID-19 misleading information policy in line with the corrected language only. We do not take enforcement action against claims that people who receive the vaccine can spread or shed the virus.
It's not clear why it took so long for Twitter to update this part of the COVID-19 misinformation policy. The new policy was introduced some time between Dec. 2 and Dec. 10, and corrected on Dec. 15 — a gap of between five and 13 days. We asked the company for an explanation of that delay, but we did not receive a response in time for publication.