Mar 5, 2015

Following a story by Digiday on Facebook, I stumbled upon the excellent report by Craig Silverman about how fake news spreads online, especially in the face of rapid-publishing media outlets. You can download it here.

I highly recommend everyone who works in journalism, digital publishing or any related fields to read the entire report. But in the interest of providing a TL;DR version, here are some insights and actionable takeaways.

In dealing with breaking unverified content on the internet, many newsrooms rush to get the eyeballs and in the process make these mistakes:

Treating journalism as an act of pointing — reporting that this rumour is what’s out there now, instead of reporting verified truths. By reporting like this, they are giving the rumour credibility via the brand, spreading misinformation. Media outlets also points to one another as their source and confirmation, so one false report sets off a chain reaction of more false reports.

Lack of follow-up on false reports — newsrooms often just grab eyeballs and move on to the next story, leaving false reports in the archive un-updated.

Declarative headlines on stories with unverified claims — headlines declare a statement or use a question like “Did….happen?” but use hedging words like “reportedly”, “claim”, “purportedly” in the story. This misleads and confuses readers, especially when people scan headlines on their news feed without clicking into the story.


Stories get more shares in the unverified rumour stage than when it has been confirmed, contributing to newsrooms’ temptation of quick and dirty reporting

Fake news articles generate more shares and social interactions than debunking articles. It took 10 large UK news websites to debunk a single fake news article. 60k shares on fake news, 60k shares total on all the debunking stories.

Strategies for newsrooms:

Newsrooms need a standard operating procedure for unconfirmed stories — workflows, choice of language/words, special website templates and headline conventions to clearly show that it is not verified.

Use SEO, clickable headlines and other traffic-generating tactics to debunk fake stories because speed and efficiency is key in fighting misinformation here. Fight fire with fire.

When a story is verified to be false, putting an “(UPDATE)” in the headline is not enough and even more damaging in spreading falsity.

Beware of cognitive challenges — inspirational and heartwarming stories e.g. sweaters for penguins are hard to debunk because people engage in “wishful thinking” and hope for it to be true. Emotions > Facts, people may not respond well.

Debunk the idea, not the person.

Be transparent in debunking — show how you found the story, why you suspect it’s not true, the methods used to find the facts.

Be very skeptical about stories and use fact-checking tools to back up.

Avoid double-hedging —

“Wednesday Apple Rumors: Apple May Be Ready to Buy Path”

Report: Possible audio tape of Michael Brown shooting”

Case study: False rumour that Durex is launching a pumpkin spice condom

Quartz contacted Durex’s PR firm, which had no comment.

Quartz broke the story saying “Durex will neither confirm or deny the pumpkin spice condom.”

Elite Daily, Uproxx, PR Newser picked up and reported.

BuzzFeed published story with confirmed denial from Durex.

Quartz updated their story.

Case study takeaways:

Update stories if the story has been proven false.

How to find rumours:

Monitor Twitter for keywords like “unverified”, “unconfirmed”, “rumour”.

Monitor RSS feeds of known rumour report sites e.g. Gawker Antiviral,, Urban Legends

Set up Google Alerts for “rumor patrol”, “unverified” etc

Follow Twitter accounts of @Hoaxolizer, @DoubtFulNews, @SkepNet

Fact-checking/rumour debunking websites & tools: