8 ways to spot fake news
During the demonetization move in November 2016, a WhatsApp forward convinced people, and even news channels, that the new Rs2,000 note came embedded with GPS trackers. The story turned out to be false.
Fake news is not a new phenomenon but social media platforms have made it much easier to spread rumours and lies.
According to a report in March by the Internet and Mobile Association of India (IAMAI), the number of internet users in India was expected to cross 450 million by June—241 million have Facebook accounts and over 200 million are on WhatsApp.
Many of these users have been mobile-first users of the internet, so they are not aware of the fake email forwards and online frauds of the desktop era. They tend to think the messages they get are genuine.
It doesn’t help that most fake news operators do not write stories that sound obviously false. They work with half-truths, turning them into believable news. “They fudge the numbers, Photoshop images, take a photo from an old source or from another country and try and sell it as statistics or a photo of something that it is not,” says Sandeep K. Shukla, head of department, computer science and engineering, at the Indian Institute of Technology, Kanpur, and project investigator at the institute’s Center for Cybersecurity and Cyber Defence of Critical Infrastructure, a centre for research on cybersecurity.
“On Facebook and WhatsApp, people share news that looks scary, spicy or confirms their bias without even reading it,” explains Shukla.
The tools for building fake stories are getting more sophisticated. A research paper by professors at Stanford University and the University of Erlangen-Nuremberg, published in 2016, demonstrated how a new video-making tool can now recreate an old news clip of a politician by changing both the expression and content. Audio advancements mean you only need 20 minutes of a voice recording to replicate it.
To counter this menace, multiple fact-checking organizations have come up, all of them trying to get to the source through filtering tools, automation and collaboration. Internet giants like Facebook and Google have tied up with fact-checking organizations to check the news they show on their feeds. In September, Facebook ran advertisements in newspapers in India, Kenya and Britain, detailing tips on spotting fake news. “The future looks positive with auto-checking of suspect stories, data mining and probabilistic reasoning,” says Shukla.
We can’t, however, depend completely on software and bots. “The algorithms and software are getting better but machines can do only as much as humans can teach them,” says Ponnurangam Kumaraguru, associate professor at the Indraprastha Institute of Information Technology, Delhi (IIIT-D). “The only way to do this is by educating the readers or consumers of content.”
The business of fake news is here to stay and Artificial Intelligence is not going to solve the issue for us, says Alexios Mantzarlis, director of the International Fact-Checking Network based in St Petersburg, US. “Language is incredibly nuanced and it’s hard to structure algorithms that automatically fact-check a claim with a high likelihood of being accurate,” he explains. Though fact-checking websites and tools will help curb the menace, the onus will ultimately lie with the reader, he adds.
So, before you react to anything you read or see on a social network, think and be very, very suspicious. Here are some tips.
Click before you share
Does the story have an eye-catching headline that tempts you to share or forward it even before you read it? “It’s important to open the link before you retweet, share or like it and go through it,” says Mantzarlis, for fake stories are designed to deceive you to maximize traffic, profit or spread of disinformation. Many a time, rumour makers use shortened URLs and catchy headlines to make it look like the source is a reputed one. Never shrug off your forwarding impulse with a “forwarded as received” disclaimer. Always read it first.
Claire Wardle, strategy and research director at First Draft, a non-profit organization working to improve practices of verification and reporting online, fake news is spread through retweets or shares without fact-checking, amplified by journalists, and pushed by trolls or loosely connected groups that are attempting to influence public opinion. “Every time we passively accept information without double-checking, or share a post, image or video before we’ve verified it, we’re adding to the noise and confusion,” she says.
Question everything you read
A study by Yale University scientists, published in September, that investigated the cognitive psychological profile of people who fall prey to fake news, found that those who apply analytical thinking to what they’re reading are less inclined to believe what they’re reading. With everything you read or view online, the first step should be to apply your brain and be willing to disbelieve. When you read something or watch a video, even if it’s coming from dependable sources online, always use your analytical powers to question the intent, the plausibility and honesty of the piece, says Shukla.
Is the URL real?
Even if the article seems to be genuine, published on a website you trust, with the same fonts and layout, it can still be fake. The internet is loaded with tools like Clone Zone that offer easy templates for people to create fake BBC or New York Times articles. Some websites use similar domain names to confuse readers, with one letter of the alphabet different in the address. A fake NYT opinion piece on WikiLeaks recently had the URL “opinion-nytimes.com while the real one is “nytimes.com/pages/opinion”. Red flags include “.com.co”, “.go.com”, “.news”, “.limo” versions of known media sites.
Install a browser plug-in
There are a few plug-ins that flag unreliable websites. BS Detector (BSdetector.tech, free), which works on Chrome and Mozilla-based browsers, checks all the links on a given webpage with unreliable sources and domains. Free Chrome plug-ins include Fake News Alert, Fake News Detector and Fake News Blocker. All these cross-check the site you’re on with a curated list of bad sites. TweetCred, developed by the IIIT-D, scores the trustworthiness of tweets by computing variables like content, user reputation, and URLs and pictures shared in the tweet. “None of these tools are perfect yet and being third-party plug-ins, they might be unsafe, but they’re helpful to an extent in pointing out problematic content,” says Kumaraguru.
Is the story a hoax?
If the first link you see to a story is from a website you haven’t heard of, and it asks you to forward it immediately, be suspicious. “Verify it by looking at other reliable sources,” says Kumaraguru. “Cross-check on websites, news channels that can verify the news you’ve just read, or websites that have a good reputation.” A quick Google search on the subject will tell you if the story has been carried by just one site or other reputed media sites too.
Cross-check with a fact-checking organization
India has a few fact-checking websites that assess trending, viral stories on Facebook, WhatsApp and Twitter and bust hoaxes. You can cross-check with sites like Altnews.in, Boomlive.in, Factchecker.in, Smxhoaxslayer.com and Check4spam.com.
International hoax-busters include the US-based Snopes (one of the first online fact-checking sites) and PolitiFact, the Brazilian fact-checking organization Boatos, the Argentinian site Chequeado, Africa Check (South Africa, Kenya, Nigeria and Senegal), machine-learning tool ClaimBuster, Faktisk in Norway, German fact-checkers WahlCheck17, and SouthAsiaCheck (Nepal).
Is this image doctored?
Reverse image search is a great first step to find out if the image you’re seeing is genuine, says Mantzarlis. Open the image in Images.google.com, Yandex.com or Tineye.com and find out where else it’s been published. Another useful plug-in for both Chrome and Firefox is RevEye, which searches for the same image across different search engines. The original one generally has the largest image size, and would have been posted first. ImgOps.com reveals information like GPS coordinates, the date of publication, the date that the picture was captured. To check how much an image has been Photoshopped, go to Izitru.com or FotoForensics.com—both can detect patterns that indicate heavy editing.
If it’s fake, flag it
If you find that a forward or an image you’ve been sent is a fake, flag it. Don’t share or promote something unless you’re certain it’s true, says Kumaraguru, but definitely talk about it if it’s fake. Tell your friends, flag it on every social network, inform fact-checking websites to stop the story from going viral.
We all need to pitch in.