The last few weeks for Amy Trott have been a marathon of labor, split between her sewing machine and the blue glow of her computer screen. She’s one of the organizers of Sew Face Masks Philadelphia — a group formed on Facebook that, through outreach on the social media site, has mobilized dozens of like-minded sewers to supply front-line workers with the most basic level of protection against the coronavirus.

Then, a week ago, just as the CDC began universal mask use, accelerating demand, Trott’s Facebook account was suspended. She spent three days in what’s commonly called Facebook jail.

“I felt like a criminal!” she said, given the implication that she’d been flagged for possibly trading in regulated goods, like firearms. More important, she said, “It completely hampered everything I was trying to do. It put everything to a halt."

Her experience is just one corner of a sprawling war over what the World Health Organization has deemed an "infodemic” of fake coronavirus news that has coursed through social media at a moment when people are more reliant on it than ever before.

In some cases, users complain that the response of such sites as Facebook has been “draconian." Their use of artificial intelligence and automated spam filters has resulted in the deletion of hundreds of posts sharing articles from legitimate news sources. And efforts to crack down on those hoarding and re-selling masks have caught up well-meaning mask-makers along with gray-market sellers. At the same time, critics say, it and other internet companies are still missing the worst of the misinformation, endangering lives in the process.

Already, at least one person has died after consuming chloroquine, interpreting advice that had spread online and was amplified with President Donald Trump’s support. A false claim that the virus spread via 5G networks led to more than a hundred incidents in Britain, the New York Times reported, including harassment of telecom workers and multiple acts of arson on wireless towers. An engineer who intentionally derailed a train in the Port of Los Angeles near the naval hospital ship Mercy said he did it to “wake people up,” because of a rumored “government takeover.” And Anthony Fauci, a doctor helping lead the White House response to the pandemic, has been assigned extra security because of threats related to online conspiracy theories.

» READ MORE: Hydroxychloroquine is not proven to work against the coronavirus. Why not just try it?

Efforts by Facebook, Google and Twitter include educational pop-ups that chase users across their websites and apps, and the prioritization of official sources such as the CDC and World Health Organization in search results.

But Brooke Binkowski, managing editor at the fact-checking site Truth or Fiction, said those efforts are nowhere near enough.

“They’re allowing this disinformation to keep spreading, and it’s been totally weaponized,” she said.

Often, social-media sites cite free speech protections preventing censorship unless that speech constitutes a true threat to public safety. "A reference is often made to shouting fire in a crowded theater. This is exponentially worse; more people’s lives are at stake. More suffering is at stake. More geopolitical concerns are at stake.”

» FAQ: Your coronavirus questions, answered.

That tolerance makes this moment particularly confusing for such people as Nicole Jochym, a medical student who has been running Sew Face Masks Philadelphia in her spare time.

“I have a lot of friends in Facebook jail right now,” she said, including one who was suspended after posting that he was 3-D printing masks to donate. Members of that mask-sewing Facebook group and many others have been experimenting with algorithm-fooling spellings such as “m@sks” and moderators now nervously debate whether to approve posts they worry might get marked as somehow violating the site’s terms.

“When [posts are] flagged in the group, you get a message that says your group is at risk: You could be disabled. So I still expect one day we will wake up and all the groups will be wiped out.”

Still, as much as users fret, this may be a time when zealous, or even overzealous, policing is warranted, said Sarah Kreps, a Cornell University professor who studies artificial intelligence and misinformation.

“I think this is a case where the public harm is such that they have to take these draconian measures to prevent people from, for example, swallowing bleach because they read that was a cure for coronavirus.”

She likened the situation to unprecedented conversations now underway about whether Americans would be willing to bargain away our right to privacy if widespread surveillance is necessary in order to end stay-at-home orders. Similarly, she thinks there’s more reason to accept potential censorship in this moment.

Kreps’ primary concern is what happens next — after the pandemic subsides. The measures that sites are taking now, she said, “would be intolerable and counter to free speech during peacetime. There is a sense in which this is unprecedented, but this should not be seen as setting precedent for the future.”