The OpenAI company has announced a new technology called "DeepFace", which allows the composition of any video with a scan of the eye mesh of a real person. The company plans to release a proprietary camera for use in this process, which eliminates the need to resort to other technological platforms for conducting compositions.
OpenAI's co-founder Sam Altman launched a project called "World", which is aimed at using eye mesh scanning to help differentiate between real images of faces and fake images created with the use of advanced artificial intelligence technologies, known as "deep fakes".
The project "World" was presented after their artificial intelligence platform for chats, ChatGPT, which allows creating fake photographs that correspond to reality with the help of artificial intelligence, was launched. Altman noted that eye mesh scanning could be used to create a "global identification document", which could only be assisted by a developed scanner called "Orb".
The main goal of the project is to solve problems related to the potential of artificial intelligence to present fake copies of video recordings of people using deep fakes. One of the main current dangers is the possibility for scammers to use deep fakes to impersonate individuals, for instance, posing as a company’s general director and demanding that their employees transfer funds to certain accounts.
