Fake Bill Keller column represents emerging form of social hoax

I suppose some congratulations are in order to the folks at WikiLeaks. They appear to have spent months planning a hoax and pounced when the right conditions presented themselves.

The result was a fake Bill Keller New York Times column that fooled journalists, including the Times’ own Nick Bilton. The column generated discussion and attracted attention to the WikiLeaks cause, which I imagine was the point.

At the same time, the hoax had a shelf life of less than 24 hours. The hoax op-ed began circulating Saturday night and on Sunday morning Keller knocked it down on Twitter:


That was enough time to make a dent in the media consciousness, but a relatively small impact given what seems to be a decent amount of time spent planning the hoax. An encouraging dynamic.

Josh Stearns put together a great Storify of how the hoax spread and how it was picked apart piece by piece after Keller tweeted it was a fake. (The Storify is embedded below.)

As is often the case with hoaxes, the trickery seems obvious once exposed. But it appears this hoax took more planning than, say, the attempt to fool people into believing that Kanye West had launched a new website.

This was more in the realm of the recent Shell Arctic hoax, which included fake websites, fake social media accounts, and even fake outrage. It was a hoax designed not just to fool, but to spread. The WikiLeaks hoax was similarly designed.

Stearn’s Storify collected a variety of revelations about the planning and execution of the hoax:

  • The domain that housed the fake column was opinion-nytimes.com. That would seem to be a red flag, as NYTimes.com is the real site. But many who subsequently shared the column used Twitter, which means the URL people clicked on was shortened. (Though it’s interesting that the WikiLeaks account, for example, did not use a shortener when it first tweeted the hoax.)
  • The hoax page layout was very similar to how the New York Times website looks, though with some alterations. For example, New York Times op-eds do not have a “tweet this” button at the top of the story. Admittedly, though, this isn’t something that will jump out at readers. The page looked real enough. And the prominent Twitter button suggests the hoaxsters saw Twitter as the best platform to help spread their work.
  • A look at the who-is record for the site shows it was registered back in March, which suggests a notable measure of forethought.
  • They also created the fake Twitter account @nytkeIler (with an uppercase “i” that appeared to be an “l”) that looked like the real thing at first glance. Many were fooled by this element of the hoax.
  • Finally, the element of timing was important. I think those involved launched it when they did because of actual comments about WikiLeaks that Keller made last week to GigaOm writer Mathew Ingram. Some of Keller’s comments were used in the fake op-ed, giving it a basis in reality. For some, it perhaps created confusion as to which came first: the GigOm post or the op-ed. Given that content flows around so much and is quoted and requoted and reblogged, the hoaxsters were smart to create this kind of confusion, and to include real words in a fake column. I also think they were wise to launch it on a Saturday night. Keller (the real one) is not a constant tweeter and chances were good that he wouldn’t be watching the network at that time, which gave the fake column a head start.

As with the Shell Arctic example, the amount of preparation and forethought places this fakery in the category of a new breed of social hoax. Note that the above elements also represent a blueprint that subsequent tricks will likely follow. So it’s also useful for journalists who want to vet information for red flags.

For tips on how to spot a hoax of this nature, have a look at my B.S. Detection presentation:

And here’s Josh Stearns’ excellent Storify about the WikiLeaks hoax:


We have made it easy to comment on posts, however we require civility and encourage full names to that end (first initial, last name is OK). Please read our guidelines here before commenting.