We live in a time of unprecedented circulation. Information-sharing is at its peak, and anything can be posted on social media and reach millions of people in seconds. This ease of communication provides many benefits, enabling real-time conversation with others, sharing emergent news, and bringing people closer digitally when physical distance is a problem. This has been especially beneficial for activist groups or other social groups who can find a sense of community in these online spaces and converse synchronously. Unfortunately, this constant stream of communication comes with drawbacks. With so much information posted every moment, there is little to no moderation, and news is often taken at face value by viewers.
In many ways, social media functions as an echo chamber. Media algorithms pair users with creators, posts, or trends that they believe users will “like” or otherwise engage with. This results in an online environment where users only view and interact with media that repeats their preconceived opinions and ideas, including “news” viewed online. Viewers who see information online that reinforces these personal beliefs, or who feel the need to respond by asserting these beliefs themselves, participate in conversations through sharing information without considering or confirming its authenticity, often unintentionally spreading misinformation in the process, while contributing to the echo chamber effect. Social media users often struggle to identify misinformation, but it becomes more challenging for users when viewing deepfakes.
The term “deepfake” refers to hyper-realistic imagery, videos, or audio recordings that have been manipulated by AI to make individuals appear as if they are saying or doing something that they never actually did, and can include face-swapping, voice cloning, or lip-synching. Deepfakes can even be used to fabricate portraits or videos of people who do not actually exist, a growing trend of deception for inhuman social media accounts to gain likes, followers, and donations. These fabricated videos can also include the circulation of humiliating, defamatory, or illegal content, including false information about political figures, malicious audio files falsified by disgruntled employees, or even fabricated pornographic images of children. This content can damage victims’ reputations and employability, especially when used to spread false narratives about others, including university staff members, students, and public figures. Politicians are especially targeted close to election seasons, and deepfakes can often be persuasive enough to sway public opinion.
With the increasing circulation of deepfakes in online media, it is becoming more challenging to determine reality from fiction, and readers’ trust in media news is wavering. Further, studies show that not only are viewers more likely to believe deepfaked videos are real, but also that most people tend to overestimate their ability to detect deepfakes, with susceptibility for deception increasing with age. However, deepfakes are not limited strictly to social media, which has become especially problematic for collegiate instructors who have not kept up with the rapid technological advancements of artificial-intelligence-made deepfake technology.
Instructors have told students for years to “research the research” when it comes to online information. The one exception has been peer-reviewed journal articles, deemed the most credible sources possible. However, even peer-reviewed journals have now become susceptible to misinformation at an alarming rate. Deepfake technology has been utilized to create fabricated images and videos to manipulate scientific publications, and research articles have also been pulled from circulation due to containing inauthentic information or misinformation, raising concerns about the credibility of these journals.
With deepfake technology just beginning to make its way into academia, it is paramount that universities respond now, rather than later, for the benefit of its students and staff. Many institutions have begun to include an “Artificial Intelligence” or “Generative AI” section on syllabi or training modules for staff to learn more about AI in general, but deepfakes are a niche element of this larger entity that require special attention. Including honor codes, clear expectations, and discussions on AI ethics in syllabi and technology centers is the bare minimum universities can do to educate students and faculty, but a much better approach is to partner with academic libraries for more thorough information literacy programs. While this subject is still being explored, both academic and public libraries are equipped to provide resources to educate people on identifying and evaluating information, a practice which is expanding to include fake news and deepfakes online.
Many universities offer library literacy modules in first-year writing or freshman seminar courses, where students are introduced to library resources and services. These learning modules offered by university libraries should be modified to include information about what deepfakes are, why they are prevalent in online spaces, and how to recognize them (when possible). By promoting enhanced awareness of deepfakes and their circulation, universities can protect both students and staff from misinformation, scams, and other dangers.
With the continuous evolution of digital spaces and the persistent spread of fake news and deepfakes, academic spaces cannot afford to remain stagnant in times of change. It is paramount that universities not only respond to these challenges, but take active measures to educate students and staff about the dangers of deepfakes.
