An Automated Way to Identify Fake Facebook Profiles Skip to main content
Utah's Foremost Platform for Undergraduate Research Presentation
2025 Abstracts

An Automated Way to Identify Fake Facebook Profiles

Author(s): Kayla Ou
Mentor(s): Xinru Page, Kirsten Chapman
Institution BYU

Autistic young adults often face unique and heightened risks on social media, stemming from a literal interpretation of platform features like Friend" requests. This tendency can lead them to overshare or trust harmful individuals, increasing their vulnerability to physical, social, and financial harm (Page et al., 2022). They may perceive a friend request as a genuine offer of friendship, which could result in dangerous offline interactions or exploitation. Research highlights that social media’s design unintentionally exacerbates these risks by not accounting for neurodiverse users, leaving them particularly susceptible to bad actors (Page et al., 2022). To address these challenges, we developed a wizard that flags potentially fake accounts. The wizard identifies suspicious profiles based on five known attributes, such as profile photo and duplicate profiles. This feature aims to reduce risks by helping autistic young adults make more informed decisions about their online connections. The tool was refined through participatory design involving autistic individuals and autism researchers, ensuring it meets the unique needs of its users. Initial findings indicate that the wizard significantly outperforms manual identification, providing faster and more accurate detection of fake profiles. This automation not only saves time but also reduces the cognitive load associated with assessing multiple profiles, making social media interactions safer and more manageable. Our next step involves conducting user testing with autistic young adults to further evaluate the tool’s effectiveness and gather feedback for improvement. This stage will provide crucial insights into how the wizard fits into real-world usage, ensuring that it continues to serve its intended purpose effectively. Full findings will be shared at the conference.