Would you be able to identify a digitally altered video if you encountered one? The challenge of distinguishing authentic content from manipulated media has become increasingly difficult. As sophisticated artificial intelligence technology for creating convincing "deepfakes" becomes more accessible, the line between reality and fabrication continues to blur. A groundbreaking digital initiative from MIT's Center for Advanced Virtuality, "In Event of Moon Disaster," aims to raise public awareness about the complex world of AI-generated media manipulation.
This thought-provoking website presents a meticulously crafted deepfake featuring U.S. President Richard M. Nixon delivering an authentic contingency speech originally written in 1969, prepared for the tragic scenario in which the Apollo 11 astronauts couldn't return from the moon. The production team collaborated with a voice actor and Respeecher, utilizing advanced deep learning techniques to generate synthetic speech. Additionally, they partnered with Canny AI to implement video dialogue replacement technology, analyzing and recreating Nixon's facial movements and lip synchronization. Through these sophisticated artificial intelligence and machine learning technologies, the resulting seven-minute film demonstrates just how convincingly realistic deepfake content can appear.
"Media misinformation represents a longstanding challenge, but amplified by deepfake technologies and the rapid dissemination of content online, it has evolved into one of the most critical issues of our time," explains D. Fox Harrell, professor of digital media and artificial intelligence at MIT and director of the MIT Center for Advanced Virtuality, part of MIT Open Learning. "Through this project—along with an educational curriculum on misinformation being developed around it—our exceptionally talented XR Creative Director Francesca Panetta is advancing one of the center's fundamental objectives: leveraging AI and virtuality technologies to support both creative expression and truth."
Complementing the film, moondisaster.org offers an extensive collection of interactive and educational resources focused on deepfake technology. Under the guidance of Panetta and Halsey Burgund, a fellow at MIT Open Documentary Lab, an interdisciplinary team comprising artists, journalists, filmmakers, designers, and computer scientists has developed a comprehensive, interactive resource platform. Here, educators and media consumers can enhance their understanding of deepfakes: their creation process and functionality; potential applications and misuses; current countermeasures being developed; and valuable teaching and learning materials.
"This alternative historical narrative demonstrates how emerging technologies can obscure truth in our environment, prompting our audience to critically examine the media they encounter daily," notes Panetta.
Launching alongside the website is a new documentary, "To Make a Deepfake," a 30-minute production by Scientific American that uses "In Event of Moon Disaster" as a starting point to explain the technology behind AI-generated media. The documentary features prominent scholars and experts discussing the current state of deepfake technology, the implications for misinformation proliferation, the distortion of our digital reality, and the future of truth in media.
The initiative receives support from the MIT Open Documentary Lab and the Mozilla Foundation, which honored "In Event of Moon Disaster" with a Creative Media Award last year. These awards align with Mozilla's mission to promote more trustworthy AI in consumer technology. The latest recipients of these awards employ art and advocacy to examine AI's impact on media and truth.
"AI has become central to consumer technology today—it curates our news, recommends potential partners, and targets us with advertisements," states J. Bob Alotta, Mozilla's vice president of global programs. "Such a powerful technology should demonstrate trustworthiness, yet often falls short. Mozilla's Creative Media Awards highlight this issue while advocating for greater privacy, transparency, and human welfare in AI development and implementation."
"In Event of Moon Disaster" was first previewed last autumn as a physical art installation at the International Documentary Film Festival Amsterdam, where it received the Special Jury Prize for Digital Storytelling. The project was subsequently selected for the 2020 Tribeca Film Festival and Cannes XR. The new website represents the project's global digital launch, making the film and associated materials freely accessible to audiences worldwide.
Recent months have witnessed a near-complete global shift to online platforms: education, entertainment, cultural institutions, political campaigns, healthcare services—all have rapidly transitioned to virtual formats. When every interaction with the world occurs through a digital lens, the ability to distinguish between authentic and manipulated media becomes more crucial than ever.
"We hope this project will encourage the public to recognize that manipulated media constitutes a significant element of our media landscape," reflects co-director Burgund. "And that through enhanced understanding and vigilance, we can all reduce our susceptibility to undue influence by such content."
