Tracing Origins: Navigating Content Authenticity in the Deepfake Era

With the wide availability of generative AI and deepfake technology, the world needs a way to identify the original sources of content and the alterations that occur throughout its existence. Every day, we are bombarded with fake images and videos of celebrities, wars, and politicians. For the last several years, hardware, software, photography, news media, and microprocessor companies have joined forces as Coalition for Content Provenance and Authenticity (C2PA) members. The goal of the C2PA is to provide hardware and software solutions to trace the origin of different types of media. Many of these same companies came together this year to sign a tech accord to combat the deceptive use of AI in the 2024 elections. As a result, support for C2PA is proliferating throughout the ecosystem, with the solution now being embedded into cameras, mobile phones, generative AI solutions, and news media reporting. However, this is just the beginning of the process. C2PA will span a multi-year journey with challenges that require deep collaboration with the security and privacy community. This is a complex cryptographic and technical solution that must be understandable by the average media consumer. From an operational perspective, the ecosystem must be robust and defend itself against potential threats from nation-state adversaries. Successful implementations must operate seamlessly at both the hardware and software layers. In addition, provenance must allow clear identification of authors while also supporting anonymity. This presentation will discuss the current state of deepfakes, the challenges that lie ahead, and how initiatives like the C2PA and other specifications are trying to address these issues.