During Hurricane Sandy's peak, Twitter was abuzz with activity, as tens of thousands of people turned to the microblogging service for alerts, updates and real-time reports and photographs of the storm.
Trouble is, not all of it was true.
Deliberate falsehoods, including images showing the Statue of Liberty engulfed in ominous clouds and sharks swimming through waterlogged suburban neighborhoods quickly spread through the service, as did word that power would be shut off for the entire city of New York and that the floor of the New York Stock Exchange had been flooded.
Twitter says it cannot possibly regulate the millions of messages on its service, and that a bit of misinformation and mischief is to be expected. But in recent years, the service has become an indispensable funnel of information in critical times for many people, especially for those who lose access to power or cable TV. For them, it can be a crucial lifeline.
Twitter has also become an invaluable tool for local government, news and relief organizations, which can monitor updates and dispatches from smartphone-wielding citizens.
"The limits of Twitter come into play during a situation like a hurricane," said John Palfrey, co-author of "Born Digital," an examination of Internet technology's impact on society. "There's great power in it, but also great power for deception."
While people are learning to accept information posted to the social Web with a large grain of salt, they may not be able to distinguish between useful updates and fake ones during a crisis or disaster, which can become dangerous. Bad information might set off an irrational decision that could lead to panic, or worse, put them in harm's way.
A search is on for promising technological solutions, but for the time being, it is up to users to police the service.
"Generally speaking, we don't mediate or moderate and monitor the site and content on the site in an event, whether the debates or a natural disaster," Rachael Horwitz, a Twitter spokeswoman, said in an interview. "Fact-checking Twitter is not scalable and not something we want to get involved with."
Twitter does take steps to validate some of the information coursing through its service. It verifies the accounts of some prominent people by placing blue checkmarks next to those users' names. It also suggests lists of reliable Twitter accounts and users to follow during big events, be they government officials or news reporters.
Ms. Horwitz compared Twitter to a self-cleaning oven, saying the company was pleased with how quickly vigilant at-home fact-checkers were able to establish which photographs and reports were not true.
Jack Stuef, a writer at BuzzFeed, exposed a Twitter user going by the moniker "ComfortablySmug," who was deliberately spreading rumors and false reports late Monday night -- one said the Coney Island Hospital was on fire -- as the frenzy online was reaching its peak. Many of the rumors were unwittingly repeated and spread by others. (Twitter said it had no plans to suspend the account.)
A Web site, IsTwitterWrong, run by Tom Phillips, an international editor at MSN, collected and analyzed popular images floating around during the hurricane and tried to determine which were real and which were not.
Twitter should not be expected to moderate all messages on its site, Mr. Palfrey said, especially during a fast-paced, critical situation like a tremendous storm.
Twitter said it did not yet have official figures for the volume of messages during the height of the storm. Topsy, a third-party service that analyzes Twitter activity clustered around a specific topic, showed as many as 550,000 posts related to the storm.
Mr. Palfrey suggested that during a major event, Twitter could pick out certain posts containing valuable nuggets of information and promote them on its site and mobile application.
Relying on users to police the streams of information could be an effective solution, said Jonathan Zittrain, a professor of Internet law at Harvard.
"Don't forget that the whole idea of a retweet was invented by users before Twitter was blessed with that functionality," he said. "The digital realm could resolve this and not break the law of physics by doing so."
He suggested an annotation, like "xTweet," to signal something as false, or even a bit of metadata that would flag a message as inaccurate, even after a posting is passed around by others.
"We could harness the viral power of Twitter to keep up with its breathless self," he said.
Correction: November 1, 2012, Thursday
This article has been revised to reflect the following correction: An earlier version of this article misspelled the surname of a BuzzFeed writer. He is Jack Stuef, not Steuf.
This article originally appeared in The New York Times.