With the rise of AI, the problems arising from it has had a wild ride, and counter measures have been sorely lacking, however, Honor has now made one of the first fully functional systems to combat AI generated contend, and especially Deepfakes.
Honor elaborates that: “HONOR’s on-device AI Deepfake Detection represents a groundbreaking solution designed to protect users from the increasingly prevalent threat of deepfakes. This innovative detection system employs advanced algorithms to meticulously analyze various elements, including pixel-level synthetic imperfections, border compositing artifacts, inter-frame continuity, consistency in face-to-ear hairstyle, and the positioning of synthetic traces”.
Unfortunately it is currently only on-device, and is a part of Honor’s new AI Agent software suite. However, this also means that no data, private or not, is shared or sent out of the device, and reduces the counter-counter measures that can be taken.
It is also a huge leap forward for IT security as Deepfake has risen to a degree where even government systems in some countries like South Korea has had images of staff removed to combat the problem.