Abstract
This technical project reflects on the lack of centralized outreach and communication tools for Contracted Individual Organizations (CIO) at the University of Virginia. Currently, CIOs rely on a fragmented combination of several platforms such as GroupMe, Instagram, and email listservs, making it difficult for students to consistently discover and engage with organizations. HoosHub was developed to unify these workflows into a single centralized web-based platform that streamlines event discovery, organization browsing, and event management.
The application is built on a React frontend and a NodeJS backend, with Amazon S3 integration to handle data storage and media management. The core feature of the application is a campus-wide event hub where CIOs can post announcements and recruit members in a more organized and discoverable format than on existing mass social media platforms like the two aforementioned. Users can only sign up for an account by using their valid @virginia.edu email, restricting the user scope to solely current members of the UVA community.
In response, the project delivers a minimalist, focused set of tools designed to make CIO announcements and events more visible and accessible to the greater student body. By providing CIOs with a global bulletin board, a refined search feature, and compact profile pages, the application consolidates the outreach process into one accessible, privacy-conscious platform exclusive to the UVA community. The result is a dedicated space where CIO visibility and student engagement can coexist without the feature-bloat and noise of broader social media platforms. The platform is designed for deployment on Heroku, a platform-as-a-service provider, and is structured to support future scaling as CIO adoption grows.
This STS paper explores current and past data privacy violations through conducting case studies on three major AI development companies: OpenAI, Anthropic, and xAI. By examining these companies under a public policy and philosophy lens, I highlight key behaviors among these companies in response to legal action and public outcry. Most notably, legal and public pressure, rather than internal ethics, are the primary driver of privacy reform across the AI industry.
By stressing gaps in legislation, especially in US legislation, the stakes of these issues due to their scale, and how adequate accountability could feasibly look, I show how the current regulatory landscape is insufficient to meet the pace and scale of AI-driven privacy violations. Consequently, these companies consistently only act on privacy violations when explicitly forced to by legal pressure. Without proactive legislative policy reform or action, the burden of privacy protection falls not on those who collect and handle data, but on those whose data is collected.
The HoosHub project and the STS paper sit at two ends of the topic of privacy. The former deals with intentionally privacy-focused design, specifically through minimal user interaction, and the latter deals with systemic privacy negligence. Building privacy into HoosHub from the ground up was straightforward due to its scale; the STS research suggests that though privacy is difficult to scale, it does not excuse larger platforms from the same responsibility. By highlighting how AI-developing businesses continue to put aside privacy for the purpose of continuing their business practices, it illustrates the impact that these violations have on their user bases and serves as a case for why privacy-conscious design must be treated as a foundational requirement rather than an afterthought.
