Deepfakes, the AI-generated fake videos, is again at the top of controversies as it was used to spoil the reputation of multiple notable personalities. The AI technology, Deepfakes can also be used to manipulate already existing content such as videos, images, and audio.
Deepfakes have wreaked havoc in federal and international politics and have impacted the public images of individuals on crucial occasions.
Multinational software giants Adobe and Microsoft have reportedly collaborated with an aim to develop new tools to detect the circulating Deepfakes and to verify the attributes and history of images and videos on the web. The government and related organizations announced their support for the new lie detector.
What Is Deepfakes?
Deepfakes is an AI-generated media that can be used to regenerate an already existing media content with new fabrications.
The challenge is when the technology is used to give form to individuals and forms that do not even actually exist in the real world, nor in the fictional world.
A journalist named Maisy Kinsley working for Bloomberg exists in the virtual world. Kinsley has a profile on LinkedIn as well as Twitter.
Another virtual avatar is Katie Johns who claims on her LinkedIn profile that she is an employee at the Center for Strategic and International Studies. The creators of this can be anyone, ranging from an individual who would have done this to misleading people or the government for a foreign spy operation, or a media group for a sting operation.
Audio notes can also be made subject to the technology of Deepfakes. The voice skins of the original sound can be used to clone the voice and fabricate a new one.
Who Created Deepfakes?
The origin of the technology can not be attributed to a single person or movement. Photo manipulation technologies have been used by mankind as ear;y as the nineteenth century.
Although the Deepfakes of the day – the viral contents of today that link Deepfakes with the credit of their formation – can be associated with Ian Goodfellow, an American computer scientist, and engineer, who is an expert in artificial neural networks.
Goodfellow invented the Generative Adversarial Networks (GAN), a network generator capable of generating plausible data. It was released in the year 2014. It was a piece of big news on the internet following its release.
The Generative Adversarial Networks (GAN) can be widely acknowledged of having two AI agents.
While
- one of the agent is to forge an image, generator
the
- other agent is to attempt to detect the forgery, discriminator
When the second agent detects forgery the other agent undergoes adaptation and makes improvements.
The initial image results after the synthesizing will look nothing like human faces. Although repeating the process many times, with feedback on the results, and the discriminator and generator both will improve.
Given enough cycles and feedback, the generator would start to produce really realistic faces of completely nonexistent celebrities.
Deepfakes were not created as a tool to promote violent content. Initially, it was used as an entertainment tool, but it also had the role of an academic tool.
Technology can’t be restricted to particular uses only as it has a broad spectrum of scopes.
Deepfakes was misused to commit cybercrimes, to run campaigns with false information to mislead people, to target a specific person, and virtually attack them, and for online fraud, and other fraudulent schemes.
The probable threats that can arise from Deepfakes are: Authentication: Distortions in ID verifications, and individual authentications, Prohibited access can be gained; Extortion, Fraud, and Reputation Risk.
What Do Deepfakes Actually Does?
Deepfakes is an advanced technology that can actually deceive people of what they see with their eyes by making it appear as convincing as possible.
Deepfakes have gained widespread attention for their potential use by creating celebrity pornographic videos, child sexual abuse material, revenge porn, fake news, bullying, hoaxes, and financial fraud. This has resulted in responses from both industry and government to detect and limit their usage
From traditional entertainment to gaming, deepfake technology has undergone an evolution to be increasingly convincing and available to the general public, permitting the disruption of the media and entertainment industries.
Deepfakes: In Popular Culture
Deepfake technology is part of the plot of One drama The Capture, the 2019 BBC.
Aloe Blacc honored his long-time collaborator Avicii, the popular figure, four years after his death by performing using deepfake technologies, their song “Wake Me Up” in English, Spanish, and Mandarin,
Deepfakes: Legal Aspect
The United States has adopted legal measures and has passed laws against the technology being misused.
Some of the states in the US have also adopted separate measures under the authority and court of the state as a combat method to face the challenges of AI technology.
They, the states are:
- Virginia
- Texas
- New York, and
- California
Other countries to do that are:
- The United Kingdom
- Canada, and
- China
How To Make Deepfakes?
Everyone from industrial researchers, academic and to amateur enthusiasts, porn producers., and visual effects studios. Governments might also be dabbling in technology, as part of their online strategies to discredit and disrupt violent extremist groups, or to make contact with targeted individuals, for example.
It is hard to make a good deepfake on just a standard computer. Most of them are created on high-end desktops with powerful graphics cards This reduces the processing time from days and weeks to hours.
Must Read:- Is Character.AI Down? How Are The Users Responding?
How Can You Detect Deepfakes?
The US researchers In 2018, discovered that deepfake faces don’t blink normally. No surprise there: the majority of images show people with their open eyes, so the algorithms never really learn about the blinking. At first, it seemed like a silver bullet for the detection procedure.
But no sooner had the research been published, than deep fakes started to blink. Such is the nature of technology : as soon as a weakness is revealed, it is fixed.
Poor-quality Deepfakes are more easier to spot. The lip-synching may be worse, or the skin tone may be patchy.
Read More:- Apple Announces New HomePod Worth $299: Check Out The Specifications
