Fake news is passé. Now, Deepfake videos are giving netizens headache
Last year, Facebook CEO Mark Zuckerberg badmouthed his company. A few days later, Barak Obama "commented" on Donald Trump. Both the videos went viral. Later, both Zuckerberg and Obama had to clarify that the videos were bogus. Welcome to the world of Deepfake videos where it's nearly impossible to distinguish between original and fake videos
Vivek Shukla 20 Sep 2019 1:14 PM GMT
Say, Mark Zuckerberg, the chief executive officer of Facebook, releases a video in which he confesses stealing personal information of its users and says that he is considering to shut Facebook down. In all probability, you would be viewing that video and sharing it too, right? Very few among us would actually think about the authenticity of that video before hitting that share button.
Those who do end up sharing such seemingly fake videos get fooled by Artificial Intelligence. Using Generative Adversarial Networks (GANs) technology, these days it's possible for anyone to make a fake video using a technique popularly known as Deepfake.
"These fake videos look so real that it's almost impossible to distinguish between the real one and the fake one," said Nitnem Singh Sodhi, a cyber expert based in Lucknow.
"These days, hundreds of videos are doing the rounds, in which a celebrity or a politician is saying something, but the voice belongs to someone else. Such videos often go viral and get shared. "There are many software available in the online market. Anyone can download the Deepfake software and create very convincing fake videos by editing photos and videos," added Nitnem, who has been working on this subject for the past eight years.
Last year, a video of Mark Zuckerberg did go viral, in which he was talking negatively about his own company. After a few days Barak Obama, former President of the USA, "commented" on current President Donald Trump. Both the videos went viral and both Zuckerberg and Obama had to clarify that the videos were bogus. These videos were made using Deepfake using Artificial Intelligence.
Who started this?
When Gaon Connection reporter asked Nitnem about the origin of this menace, he said: "It is difficult to pinpoint one person. If people want to stay away from such videos, they should not share such videos without authenticating them. Citizens will have to behave responsibly."
According to a report, Korean company Samsung invented this Deepfake software wherein a single picture could be converted into a full-fledged video clip.
These days, the Deepfake software is used during elections to spread fake news and videos. This is done to malign the image of rival candidates. In the case of celebrities, their fake sex tapes or "porn videos" are released to mock them or to ruin their marriages. These videos often go viral on YouTube, WhatsApp, porn sites or Vimeo.
"Anyone can make Deepfake videos. People are using this software to malign images of politicians, including the prominent ones like Prime Minister Narendra Modi or Rahul Gandhi. This is a modified version of fake news," said Chatak Bajpai, a cybersecurity expert.
This is how you can identify Deepfake videos
Dr Mukul Srivastava, Dean of Department of Journalism and Mass Communication in the University of Lucknow, told Gaon Connection that Google and Yandex have provided reverse image service, wherein these search engines search for images that are exact copies of the source image or that differ in minor ways from the source image. Using this, one can upload a suspect image and then look for its source image's date of uploading.
He has written many articles on cyber-crime and has been spreading awareness through Facebook.
While explaining the pros and cons of Deepfake videos, he told Gaon Connection that Amnesty International, a London-based non-profit organization, which works to protect people from social abuses, has teamed up with YouTube and has launched YouTube Viewer to verify videos uploaded and its background.
"In order to authenticate the video, one needs to re-play the video again and again. One should always look for specific details like posters, banners, number plate of vehicles, landmarks, clothes of people in the video, their spoken language etc. All this should be considered before falling into the trap. One should think why a particular video was sent to them. Was it shared to spread awareness, educate or for provocation?"
Cyber-crime has been on a rise
According to security software firm Symantec, in 2017, India was ranked third globally, behind the US and China, as a source of malicious activity. According to the National Crime Records Bureau (NCRB), over 966 cyber-crimes were reported in 2010, which rose to 12,317 in 2016. In other words, there has been an increase of 1,175% in cyber-crimes over the past six years.
Creating Deepfake video is a criminal offence. Cyber-crimes have costed India Rs 24,630 crore in 2013 alone, according to a report commissioned by the Delhi High Court.
Cyber security crimes, such as phishing, scanning, introducing malicious code, website intrusion and denial of service, rose to 76% over the past five years --from 28,127 in 2011 to 49,455 in 2015. While 28,127 cyber-crimes were reported in 2011, they rose to 49,455 in 2016. Financial fraud, stalking, online identity theft, forgery, information piracy have been categorized as the common cyber-crimes these days.
Also Read: Dear Amit Shah, Digital India is great, but please deal with cyber-crime too
Also Read: Mobile-addict kids: It's a ticking time bomb