DARPA prepares new anti-deepfake technology
The US Defense Advanced Research Projects Agency has said its Forensic Semantics program is ready for use by industry and academia.
Deepfakes are a growing problem: Whether it’s celebrities trapped in deepfake porn, advanced threat actors using deepfakes of politicians to sow political discord, or scammers posing as friends and family, deepfakes are a widespread threat.
However, the US Defense Advanced Research Projects Agency has been working on a solution to the problem and is ready to bring its work to a wider audience.
DARPA’s Semantic Forensics program, also known as SemaFor, is designed to help people and organizations detect deepfakes.
“With SemaFor in its final phase, DARPA investments have systematically reduced development risks, paving the way for a new era of defenses against the growing threat of deepfakes,” DARPA said in a statement.
“Now, the agency is calling on the broader community, including commercial industry and academia doing research in this space, to take advantage of these investments.”
DARPA is doing two things with its SemaFor program to help advocate for the broader community. The first is the launch of an open source analytical catalog developed by the program, which will be shared with researchers and industry.
The second effort is the DARPA AI Forensic Research Open Research Challenge Evaluation, or AI FORCE. Its goal is to develop machine learning models that can detect AI-generated images through a series of “mini-challenges” where participants must choose between fully synthetic images, authentic images, and images once retouched by other methods.
According to Dr. Wil Corvey, director of DARPA’s Forensic Semantics program, it’s time for the program to grow.
“Our investments have seeded a space of opportunity that is timely, necessary and poised for growth,” Corvey said.
“With the help of industry and academia around the world, the Semantic Forensics program is ready to share what we have started to strengthen the ecosystem necessary to defend authenticity in a digital world.”
