People can no longer tell the difference between a real human face and a fake one created by AI, a new studysays. The concept of creating fake identities is not new, but it has been taken to new extremes with the rise of digital 'Deep Fakes.' AI and machine learning are being used to generate digital humans and they appear to be getting better by the hour.
AI programs that pretend to be human have the potential to take on many roles currently occupied by human workers, with them viewed as perfect and tireless workers for digital bank services, customer services, call centers, sales, and support. Profile avatars created by AI are also heavily used to create fake profiles and have been linked to fraud and cybersecurity threats.
Related: This Creepy New AI Is Way Too Good At Pretending To Be Human
Researchers of the University of Lancaster concluded that humans can no longer distinguish between a face generated by artificial intelligence and a real human face. The team says this represents an urgent risk that needs to be addressed by creating safeguards to protect the public. The researchers added that AI-synthesized text, audio, image, and video have already been used for various and nefarious purposes, including fraud and propaganda.
The researchers explained that the artificial faces were made using StyleGAN2, a generative adversarial network originally created by Nvidia researchers in 2018. The experiment contained more than 500 participants who were asked to distinguish synthesized faces from real ones, and asked to determine a level of trust that the faces impressed on them. The results showed that AI images are not just photo-realistic, but almost indistinguishable. Even more concerning, the study found that people
Read more on screenrant.com