It is now out that social media platforms, Snapchat and Tik Tok are on a race to be the first to add Deep Fake Technology to their applications.
Snapchat appears to be ahead in the race after quietly acquiring AI Factory, another Ukraine-based computer vision company at a price reportedly amounting to $166 million. According to TechCrunch,Tik Tok on the other hand, is said to be working on a direct deepfake-style tool with features similar to the Chinese video editing app ZAO.
What is Deep Fake Technology?
Deep Fake is a tool using Artificial Intelligence (AI) to create or modify video content in such a way that something that did not actually happen may be presented as alternative content.
The term deep fake was actually derived from a Reddit user who went by the name “deepfakes,” when in December 2017, he had used deep learning technology to replace faces of characters appearing in pronographic video clips with faces of well known celebrities. The term may refer to both the deep learning technologies and the fake videos created using the technology.
About the Deep Fake Technology Acquired by Snapchat
Although the news about the AI Factory acquisition has been confirmed by Snapchat, the latter still has to come out with a press release to officially confirm the news reported by Ukraine’s AIN.UA publication.
AI Factory was founded by Ukraine-based Victor Shaburov, Greg Tkachenko and Alexander Mashrabov. Shaburov, who is also the founder of another computer vision firm Looksery, was also Snapchat’s Director of Engineering. He left Snap in May 2018 to lead the founding of AI Factory with Tkachenko and Mashrabov. As of the moment AI Factory founders have not yet commented about this report.
Still, it is widely known that Snapchat had worked with the AI Factory in developing its recently activated “Cameo” feature. The “Cameo” tool allows Snapchat users to paste their faces as overlays on a selection of usable pre-made scenes.
TechCrunch Voices Concerns Over the Addition of Deep Fake Technology to the Social Media Apps
Although TechCrunch reports that the deep fake tools produce content that still look rough around the edges, it is likely that improvements will come around real fast. The danger however is that up to what point will the deep fake tool be used for fun, where fun is just for online entertainment.
Techcrunch raises concerns that the technology can also be used for nefarious or criminal purposes. The potential for doing harm is great even if a video is later proven as fabricated, since depicting a person, let us say a politician or celebrity in a compromising situation or position is enough to damage reputations.
As with the deep fake tool being worked on by Tik Tok, greater concerns are being expressed since the tool involves uploading of biometric data. It is reported to be similar to the version of the Chinese app ‘Douyin, for which the purpose is to provide the Chinese Government with data to use for identification purposes.