Why We Must Provide Our Own Security in the Wild West of Deepfakes?

Everyone everywhere is as familiar with the Wild West as they are with Wild Capitalism. Some know these precarious lands from history books, but most have learned about them in Hollywood Westerns. Although there is a sheriff in town, everyone must provide their own security. Danger could come from any direction. A knock on your door could be violent felons wanting your horses, or savage Native Americans whose lands you are after. Also, psychopathic killers and perverts run rampant on those vast desolate mountains and hills. Neither homes nor trails are safe. You can lay down the law so long as you are tough enough to bring someone to their knees. That is why no-one walks around without a pistol or a rifle. Even women and children know how to use a gun. It is assumed that everyone should always be on their guard in the Wild West.

That is the frightful and perilous story of the foundation of the United States, which is regarded as the cradle of technology. Threats driven by the supreme civilization of the era moved towards the unspoiled lands from the Atlantic Coast to the Midwest and West. California’s Silicon Valley on the West Coast of the United States is the destination of this treacherous journey. Globalization will never stop spreading Wild Capitalism to other continents by using the technology weapon. The method is the same in the Wild West and Wild Capitalism alike. The one who holds the weapon uses it to his own advantage. Civilization develops the weapon first, then abuses its power for its brutal objectives. However, it does not do it openly. It always seeks to achieve its objectives secretly and through deception. The power elites are a global extension of this mighty adventure as shown in the example of Operation Gladio, a “deep state organization” that, according to recent claims, was unearthed in Italy.

Technology develops weapons, but it is up to you to protect yourself

The depth of the danger to humanity poses a great threat. And now, this depth is completing its evolution of digitalization with a technology developed by using Artificial Intelligence (AI). Consequently, the cycle of using all of the earth’s resources and reserves for one’s own global interests has reached a new phase. Synthetic media that is created with deepfake sound and images have added a new dimension to these depths. The power to divide and conquer societies exists on a greater scale and has gained more speed. Just as in the Wild West, there is no-one to protect us from the perils of deepfakes. It is not known when, how, or from where the assault will come. “You can preserve your perception of reality only with your intelligence and mind. You must protect yourself with the technology weapons you can muster,” we are warned by Globalization.

World Economic Forum (WEF), one of the most important platforms from which current globalization messages are conveyed to the public, discussed deepfakes at the Davos Summit in February. As we briefly mentioned in a recent article, the Davos Summit hosted Hao Li, Professor of Computer Science at the University of Southern California, who is described as the world’s best deepfake developer by the MIT Technology Review. Pinscreen founder and CEO Hao Li and a team from the company demonstrated how participants could experience real-time face swaps on a Virtual Mirror developed by them. But the most important take away the clear message they conveyed: “You are responsible for your own security against deepfakes.”

Cheap fakes herald the dawn of the deepfake era

Today, as an indispensable part of our lives, digital media has grown into an incurable addiction for modern society. More than 1.2 trillion digital images and videos are captured annually in the world. This figure is increasing by 10% each year with more than 2.7 billion people capturing 85% of the images on smartphones. Increasingly, these digital recordings are being manipulated, and there are a higher number of digital processes on these digital audios, photos, and videos. Possibilities in audio and visual manipulation have moved beyond the expectations of experts, thanks to the image and audio processing technologies that are developed with deep learning algorithms through AI-based artificial neural networks. The confidence in audio and visual truth has gradually waned with the speed of growth and the spread of deep synthetic media. Generative Adversarial Networks (GAN), which demand high processing and graphic card capacity, and the other machine learning-based deepfake technol ogies still have not reached a high enough production capacity to occupy the public domain. But, cheap fakes—entertaining face swap apps on smartphones—or more elementary methods, such as speeding up and slowing down for the purpose of disinformation and online fraud, are used at a sufficient level to cause uproar and concern. The number of productions is rising also in deepvoice and deepfake media, which are becoming increasingly hard to distinguish from the originals. This will inevitably result in a “break from reality” that is sure to disrupt the processes of perception and decision-making for businesses and society.

Will it ever be possible to detect deepfakes in real time?

Truepic does not offer users the absolute scanning experience that Zemana Deepware Scanner offers. Nevertheless, it describes itself as a photo and video verifying platform that offers services to journalists, NGOs, public institutions, and companies. “It will never be possible to detect and stop deepfake synthetic media in real time at the root,” asserted Mounir Ibrahim, Truepic’s Vice President, and he also made the aforementioned points in an article on the WEF website about a month after the Davos Summit. There are a number of reasons for this, one of which is that as digital media content goes around the internet, almost all of the metadata, the essential information that defines it, gets lost, destroyed or changed.

Upon scanning the detection system of an image, it will be impossible to recreate the lost metadata. Thus, details such as the original date, time, and location of the image will probably remain unidentified. Also, nearly all digital images are instantly compressed a nd resized immediately after being shared on the internet. At times, this may be an effort to mislead the content viewer. Once they are uploaded or transferred, recompression and resizing of the images make it very hard to detect any pixel manipulation. Even if an automatic or machine learning-based detection model is registered and becomes available for use, cyber-hackers would develop other methods to remain undetectable.

The source of the outbreak: social media

What makes it even harder to detect fake synthetic media is the social media that spreads any content, regardless of whether it is real or fake, within seconds throughout the world. People with malicious intent can instantly plug fake content on social media platforms. Therefore, it may be too late to halt the spread of fake content even with a proven scanning model. Lack of awareness or fanaticism will speed up the viral spread of disinformation and make objective decision-making burdensome.

At the World Economic Forum, Truepic’s Vice President claimed that it would never be possible to detect and stop fake synthetic content before it spreads throughout the online world. The solution he puts forward is to establish their authenticity based on provenance (source, root, origin), instead of trying to decide whether digital content is fake or not. Are these remarks a premarketing attempt for the new technology that the company seeks to develop? Or, despite all of the aforementioned hurdles, is it possible to quickly detect and stop fake synthetic audio and images at the root before they spread? Time will tell.

The results of the Deepfake Detection Challenge (DFDC) are in, with the expectation that a deepfake detection model will be developed for the online world. The outcome has not been officially confirmed by the organizers, and there has been some delay with the announced dates, most likely due to the pandemic. Next is the deepvoice competition initiated by Google. Sincere or not, the rulers of the online world seem to be looking for a technological solution to the deepfake nightmare that they have created. However, even if this possible, it will take a long time.

Deep civilization is now taking over people’s minds with this technology weapon. As it was in the Wild West, it is still up to us to protect ourselves against vicious attacks.