A new tech standard aims to combat deepfakes

The Coalition for Content Provenance and Authenticity (C2PA) announced on Wednesday that it had partnered with tech giants including Microsoft, Intel, and Adobe to launch a new standard aimed at impeding the rapid spread of deepfakes.

The first-of-its-kind specification grants content creators a means to develop what the organization calls “tamper-evident” media. It will work by providing platforms a way to define who created and changed information associated with images, videos, documents, and other assets, as well as identify evidence of manipulation. Andy Parsons, Senior Director of the Content Authenticity Initiative at Adobe, told The Record, “The C2PA expects its open standard to be broadly adopted across the content ecosystem – by device makers, news organizations, software and platform companies."

In a live Q&A on Wednesday hosted by C2PA, lawmakers stressed the urgency around stopping the spread of deepfakes, which Sen. Robert Portman (R., OH) described as “hyper-realistic content depicting events that actually did not occur.” 

"Deepfakes, which are unfortunately sometimes used against unsuspecting women in the form of pornography, are causing great harm in our communities,” Portman warned. The specification technology gained the support of Portman as well as other government officials for its ability to collect and communicate the history of digital media content, he said.

Intel clarified that the specification alone will not completely rid the internet of deepfakes. Jennifer Foss, Communications Manager for Intel Corporation, told the record, “Provenance information is just one important piece that can be used in conjunction with many other technologies including deep fake detection technologies.” 

Sen. Gary Peters (D., MI) said at the Wednesday panel that he considers deepfakes “a critical national security issue,” referring to the manipulation of media content (images, videos, audio files, and documents) through artificial intelligence. The escalation of this tactic poses serious threats to consumers, including the spread of disinformation, weakened trust between users’ and companies, and a means to target women, the lawmakers said.

C2PA launched last February, and its partners include the BBC, The New York Times, Twitter, Nikon, Akamai, Fastly, and others.