Twitter Needs to Deal With Misinformation. Here's How

The social network needs a better way to deal with errors and corrections.

Over at Poynter, Jeff Sonderman flags a site called Let Me Tweet That For You, which does exactly what its name suggests:

It’s pretty simple—you type in a Twitter username and a message, and it generates a realistic-looking image of a tweet from that person. It even adds fake retweet and favorite counts to lend some more credibility.

The tool is kind of fun to play with, until you realize its potential to spread havoc and misinformation. Here's a small sample of the false tweets Web users have been making for other people:

While Let Me Tweet That For You has actually been around for some time , it resurfaces a perennial issue: How do you correct something as fleeting as a tweet?

Zeynep Tufekci is a fellow at Princeton University's Center for Information Technology Policy. Earlier this month, Tufekci tweeted what she thought was the new Pope's Twitter handle. It turned out that the username was a fake. Although Tufekci corrected herself immediately after discovering the mistake, it didn't stop people from seeing her older, incorrect tweet.

In a blog post later, Tufekci called on Twitter to create a feature that would alert innocent users to misinformation. Her suggestion? Allow the creator of the original, mistaken tweet to issue the offending tweet again, but this time with a big "REDACTED" or "ERROR" sign on it.*

As a way to promote transparency and accountability among users, this isn't a bad idea. But as Tufekci points out, there's also no guarantee that everyone will see the second tweet with the correction appended. Nor would the system do anything to modify her original, mistaken tweet, which is still living in cyberspace (she didn't delete it so that there would be a record both of the error and the correction).

Given that some people almost certainly saw just the wrong information and not the correction, I'd suggest an addition to Tufekci's idea—a feature that:

  • Lets users mark their own tweets as incorrect after the fact, much in the way that users are able to mark their tweets as "favorites" now; that then
  • Flags the content publicly with a colored tab; and
  • Alerts anyone who clicks "retweet" that the tweet has been marked as incorrect by the original user.

As a result, a tweet that has been marked incorrect would appear in people's timelines like this:

And if someone tries to retweet the incorrect tweet, they'd see something like this:

Given the state of Twitter's current technology behind favoriting and retweeting, this wouldn't seem like a major lift.

*Update: Tufekci writes in with a clarification -- the "ERROR" sign in her solution would also be applied to the original (incorrect) tweet and not only to the new (corrected) tweet.