I woke up Sunday morning to a tempest on Twitter. Apparently, Bill Keller of the New York Times had published an op-ed defending Wikileaks. It was retweeted by none other than Jeff Jarvis, Dan Gillmor and Nick Bilton–some of the top media minds in the country–until, not too many hours later, it was discovered to have been entirely fabricated.
Keller repudiated the post (in ALL CAPS), Jarvis, Bilton and others rapidly apologized for having been taken in, and Harvard Berkman fellow Zeynep Tufekci, NSA whistleblower Thomas Drake, Reuters columnist Anthony DeRosa and dozens of others immediately began sleuthing to determine what had happened. Tufekci discovered a fake New York Times domain name registered on March 30, tweeting “This was in the works for a while.” Later, she added that “Whoever did this also had
We’ve seen similar dynamics many times before. Nearly a year ago, rumors of Muammar Gaddafi’s death began trending on Twitter before being retracted. Numerous celebrities–including Brad Pitt, Eddie Murphy and Robert Pattinson–are regularly the targets of death hoaxes.
Here’s a timeline of the Gaddafi death rumor of August 21, 2011, courtesy of Topsy. (Gaddafi actually died in October of that year):
Of course, it’s not only politicians, journalists and entertainers who are vulnerable to confirmation bias and the accelerating effects of social media. Businesses are as well. Snopes.com is rich with examples of urban legends that have sparked public imagination, dread and disgust. Did you know, for example, that bat excrement is a secret ingredient in mascara? That Lucky Strike cigarettes were so named because some packs contained marijuana cigarettes? Right, because they’re not true. We can all think of examples–from last week, from childhood. Thanks to the tireless folks at Snopes, and now others on the web, they eventually get corrected.
So, does the Internet make disinformation worse or better? Yes. As we see from yesterday’s New York Times kerfuffle, not only can you spread disinformation quickly via social media, but you can detect it too.
Says Rishab Ghosh, Chief Scientist, Topsy, “In social media you can track where a message came from, see who the first person who tweeted it out was and whether they are reliable. You don’t have a guarantee of whether it’s true or not, but you can see how it started.”
So, in that spirit, here are three simple things Ghosh recommends you do to identify disinformation on the web:
- Check the URL. The mistake Jarvis, Bilton and others made before retweeting the erroneous NYT op-ed was to forget to check the URL. Had they done so, they would have recognized that is wasn’t legitimate. The fake column began “http://www.opinion-nytimes.com…” while real New York Times Op-Eds look like this: “http://www.nytimes.com/pages/opinion/index.html”.
- Check the story independently against the real source. The story did not appear on the New York Times Opinion page.
- Check the Twitter account. The story was tweeted by a fake account attributed to Bill Keller. At first glance, it looked like Keller’s actual account (@NYTKeller) but turned out to be, according to Nick Bilton, spelled with an uppercase “I” and a lowercase “l”. Tricky, tricky.
To be fair, this was a pretty elaborate fake, and it took in, albeit briefly, some of the most sophisticated media folks in the country. So the lesson here is this: any of us, no matter how smart, is susceptible.
Disinformation is a real business risk, and, like other sorts of black-hat tactics, is becoming more sophisticated all the time. We are seeing promising advances in influence detection algorithm technologies, and soon technologies will emerge that will use machines to help flag the characteristics of reliable versus unreliable content. Until then, a healthy amount of skepticism and a few spot checks are in order.
I’d love your feedback, tips and tricks for identifying false content. Please leave your ideas in the comments, or if you post elsewhere, I’ll link back to you.