Generative AI Safety
The Battle Against Digital Forgery: Safeguarding Content in the AI Era
In this segment of our ongoing discussion about the threats posed by digital forgery, we focus on the current landscape of deep fake technology. We explore statistics that illuminate the prevalence of these digital deceptions and delve into notable case studies where deep fakes and unauthorized uses of likenesses have had significant impacts.
The Current Landscape
Statistics on Digital Forgery
The advent of deep fake technology has brought about a seismic shift in digital content creation, making it increasingly difficult to distinguish between what is real and what is fabricated. According to a report by Deeptrace Labs, the number of deep fake videos online nearly doubled to 14,678 in just nine months, from December 2018 to September 2019. This explosive growth is indicative of the technology's accessibility and the ease with which it can be used to create convincing fake content.
More recently, a study conducted in 2022 by Sensity AI noted that over 90% of the deep fake videos circulating on the internet are used for creating non-consensual adult content, exploiting the likenesses of unsuspecting individuals, primarily women. These statistics not only highlight the prevalence of deep fakes but also underscore the harmful uses to which they are often put, necessitating robust mechanisms for content authentication and protection.
Case Studies: Notable Incidents Involving Deep Fakes
1. Politics and Public Figures
The Pelosi Video: One of the more publicized incidents involved a manipulated video of Nancy Pelosi, the Speaker of the U.S. House of Representatives, which was slowed down to make her appear impaired. This video, which was spread across social media in 2019, highlighted the potential of deep fakes to influence public opinion and tarnish reputations.
Manipulation in International Politics: In 2020, a deep fake video of the President of Gabon on New Year's address was suspected to be a deep fake amidst rumors of his death. This incident caused significant political unrest and doubt among the citizens and international observers.
2. Entertainment and Media
The Lucasfilm Response: After a deep fake video featuring a young Luke Skywalker went viral, Lucasfilm hired the creator of the deep fake to work on its visual effects team. This move by a major film studio in 2021 acknowledged the dual-edged nature of deep fake technology—its potential for creative use in entertainment, alongside its capacity for harm.
3. Legal and Personal Implications
Unauthorized Celebrity Videos: Celebrities have often been the target of deep fakes, with their likenesses being used without consent. A striking case involved a series of deep fake videos of Tom Cruise in 2021, which were so realistic that they raised alarms about the potential for misuse in scenarios involving less benign intentions.
The landscape of digital forgery, particularly through deep fakes, presents a clear and present danger to individuals, institutions, and the integrity of public discourse. The statistics and cases discussed illuminate the critical need for advancements in content authentication technologies. As we continue to grapple with the implications of this evolving threat, the development of legal, technological, and societal safeguards will be paramount in the battle against digital forgery.