In exciting news, researchers at Facebook have revealed new technology that allows them to track deepfakes to the source. They reportedly developed an AI software that could be trained to detect deepfakes from still images or video frames, and even identify which software was used to create the fake, “no matter how novel the technique.”
Facebook officially banned deepfakes in January of 2020, but this ban proved hard to enforce. With new technologies such as the one in development, deepfake misinformation will likely slow to a halt, as new methods of discovery and identification will be usable.
For more information about the program, see this article by CNBC