The Pandemic of Fake News
During the lockdown there has been a marked increase in the incidence of fake news and its viral spread. According to a research falsehood diffused significantly farther, faster, deeper, and more broadly than the truth in all categories of information, and the effects were more pronounced for false political news than for false news about terrorism, natural disasters, science, urban legends, or financial information. False news was more novel than true news, which suggests that people were more likely to share novel information. Whereas false stories inspired fear, disgust, and surprise in replies, true stories inspired anticipation, sadness, joy, and trust. Contrary to conventional wisdom, robots accelerated the spread of true and false news at the same rate, implying that false news spreads more than the truth because humans, not robots, are more likely to spread it.
As case in point, even a celebrity like Amitabh Bachchan was not spared.
What is Fake News?
Fake or false news can be categorized in following ways:
- Fake Article. The entire article is written with the intent of spreading false information.
- Embellished Article. A real article or fact is enhanced and embellished to add fake information to it.
- Fake Photo/Fake Video. A photo or video is enhanced and modified to add fake information to it.
- Older Photo/Video or out of context photo/video. An older photo/video or a photo/video that’s from another country or another context is inserted with the intent of spreading false information.
The common intent of fake news can be:
- Communal, news written with the intention of inciting religious and communal passion
- Violence, Criminal and sexual.
- Integrity and Authenticity
- Questionable Practices and Rituals (Occult, Blackmagic)
- Bullying and Terrorism
- Misinformation (bank shutdown, currency devaluation, economy)
- False Health Advisories, we see an uptick in these during the Covid-19 pandemic
- Spreading false information about competition
- Political misinformation
- Satire
- Propaganda
- Personal Defamation
The Human, Economic and Psychological Damage from Fake News
Some estimates put the damage from Fake news in the range of 80 billion annually.
- Fake News Stock Market damage – Fake news has caused losses of billions of dollars in company stocks. The damage last year alone was $39 billion or .05 percent of the markets value [1]. A letter purportedly sent by Larry Fink, CEO of Blackrock caused investors to panic and dump coal stocks. [2]
- Health scares – fake news promotes the sale of questionable medical supplies and result in medical scares from visits to ER’s.
- Destroying government machinery and loss of life – fake news has resulted in loss of confidence in government machinery and in cases of violence that have resulted in rioting and loss of life.
- Politics – there have been innumerable news articles and posts on how fake news has resulted in election decision making in favor of the less desirable candidate.
- Loss of sales due to fake propaganda, corporate espionage and loss of trust in brands – Fake news has resulted in loss of sales and goodwill. [3] McDonalds has been targeted by multiple instances of fake news and has lost customers each time. In addition they have had to spend money each time on advertising to promote facts.
- Personality disorder in humans who read fake news analysis – Companies that manually curate news to figure out real from fake have high attrition. Employees get depressed from the psychological impact of reading fake news day in and day out.
Fake News Pandemic – How do We Tackle it and Mitigate Reputation Risk?
There are various ways social media platforms, law enforcement agencies, Governments, political parties, companies and individuals celebrities combat fake news pandemic. The tools available can be separated into three categories:
- an increase in the use of human editors by the social media and publishing platforms. Given the volume of information being disseminated, it is practically inhuman to cover all the material being published
- crowdsourcing initiatives being used by various initiative still not a viable solution of the joint forces to combat the evil forces of fake news; and
- technological or algorithmic solutions
We would like to discuss on how we use our algos to assist in mitigating the reputation risks for different stakeholders affected by fake news
Common-sense techniques to find instances of fake news
Text Analytics
- Subject detection, scan the article to figure out the theme and subject.
- Use of provocative keywords, fake news usually contains provocative keywords.
- Sense of urgency – Creating a sense of urgency to ask the reader to take action now by clicking on a link.
- Spelling and grammar mistakes.
- Creator credibility analysis, is this website credible, is this handle a new handle that hasn’t posted before
- Links to websites that contain spyware/botware or other paid news
- How fake news spreads (not spread by influencers, people with large number of followers), spread by accounts that habitually post controversial content. Truth spreads slower typically.
- Does the social media account come across as a Bot. Bots have typical characteristics such as account age, account photo and posting only articles and forwarded content. Bots spread fake news faster than individuals.
Image Analytics
Fake news shows a tendency to contain graphic images that help drive home the point of the article. The following techniques help in analyzing the image and hence determining if the article is real or fake.
- Reverse Image search, we search for images using a reverse search. This gives us context on when the image was first posted, what the original image looked like and what were the changes
- Image authenticity, an image that diverges significantly from the original or shows signs of doctoring is probably a fake. Minor alterations such as brightening and color correction are acceptable changes to the image.
- Image age, if the image is old then its irrelevant and is used to spread fake news
- Image geotags, if the geotags on the image differ from the location mentioned in the article then the image is fake and the article is fake.
Video Analytics
Similar to images, articles contain videos that are provocative to drive home the point the article is trying to make. We use the following techniques to analyze videos
- Video authenticity, a video that diverges significantly from the original or shows signs of doctoring is probably a fake. Minor alterations such as brightening and color correction are acceptable changes to the video. These videos undergo significant modifications in the post production phase.
- TikTok analysis, TikTok is a growing source of modified videos.
- Video age, if the video is old then its irrelevant and is used to spread fake news
- Video geotags, if the geotags on the video differ from the location mentioned in the article then the video is fake and the article is fake
Article Analytics
A wholistic analysis of the article as a whole gives rich insights into the authenticity of the article. Some analysis that we perform here are
- Author check, fake news generally does not provide the author name or comes from unknown authors that lack credibility.
- Text to advertising ratio, fake news generally has a higher ratio of advertising or links as compared to text
- Article layout, a badly laid out article on a website is a good indicator of the presence of fake news
Be Safe! Spread Genuine!
Footnotes
[1] Fake News Creates Real Losses https://www.institutionalinvestor.com/article/b1j2ttw22xf7n6/Fake-News-Creates-Real-Losses
[2] The Fake Larry Fink Letter That Duped Reporters https://www.institutionalinvestor.com/article/b1cqj0xmn5ds9t/The-Fake-Larry-Fink-Letter-That-Duped-Reporters
[3] French Fry Grease in McDonald’s Coffee https://www.snopes.com/fact-check/mcdonalds-french-fry-grease/
