Being Tom Cruise Is Easy
Deepfake technology may be destroying society one celebrity at a time
Deepfake technology, machine learning software that anyone can use to create false images and videos of people and events, has evolved so much since its rudimentary introduction in 2017 that most humans (and computers) can now be easily tricked into believing its true. It will rapidly emerge as the most significant fake news threat to modern society.
Just ask Tom Cruise.
Based on an analysis from the excellent News Literacy Project, a recently created TicToc account, @deeptomcruise, posted a series of three videos in late February that demonstrate the shocking sophistication of deepfake technology. In each of the videos, the technology appears to have seamlessly mapped the face of Tom Cruise onto that of an actor performing some act. Though several indicators, including the name of the TicToc account and some hyperbolic impersonation of well-known Cruise mannerisms, point to the video being fake, the visual effects look flawless to the human eye and would easily fool anyone.
Like many new technologies, the potential of deep fakes has likely been tested most in the online porn industry. The Guardian reported that 96% of online deep fake videos found in September 2019 by one firm were pornography, with a staggering 99% fantasizing sex with female celebrities. Weaponizing new technology to attack women is sadly common, as this article from Ms. Magazine details, and it usually serves as an implicit gateway for broader malfeasance. That evolution beyond its misogynistic roots is what we are currently witnessing with deepfake technology.
Since fake news is based on subverting facts just enough to make desired fantasies plausible, deepfakes promise to be effective weapons of future misinformation campaigns. A simple video search can illustrate a variety of examples, usually case studies and “what ifs” to help expose some of the frightening misuse scenarios that the technology empowers. In the same way that fake news purveyors attempt to manipulate people into sympathizing with an unsubstantiated viewpoint or belief, viral deepfakes can quickly present difficult to disprove “evidence” that discredits more trustworthy information sources and undermine the reputations of public figures (like Tom Cruise).
Lesson Learned: Seeing is no longer believing.
According to this report from The Brookings Institution, technology advances have made distinguishing between a deepfake and a real image or video in someone’s social media feed nearly impossible. Potential victims include not only celebrities or other recognizable public figures like politicians. Malicious actors can also target businesses by leveraging real executive appearances and media to create content that manipulates stock prices, sabotages reputation, and sways regulatory actions.
Key to combatting the propagation of misinformation via deepfakes is by pausing before clicking that “Share” button and taking a few moments to assess the standard potential indicators of fake news: source, history, and social media footprint. In application to the viral Tom Cruise TicTocs, it’s difficult to attribute the origination source to the actual Tom Cruise because no other verified online connection either through previously known social media accounts or from the actor’s web site, but new accounts with very limited history should always be treated cautiously. Then, patiently and politely questioning it on social media can help inform others to challenge whether their own personal biases are driving their decision to either accept or reject the underlying fake news.
