DARPA’s deepfake detection team is looking for partners to use its tools, shape policy

The Defense Advanced Research Projects Agency has amassed a trove of algorithms it has been developing for almost a decade to detect manipulated media and is pursuing their uptake by industry and federal agencies, as demand is expected to surge with the advent of generative AI.

“The team is currently focusing on outreach to USG transition partners and making toolset available to researchers,” Wil Corvey, the program manager for DARPA’s Semantic Forensics, or SemaFor, program told Inside AI Policy by...