Guess, AndyColes, Daniel T.2025-08-042025-08-042025-04-03https://theses-dissertations.princeton.edu/handle/88435/dsp01t148fm58vThis thesis investigates the impact of content moderation changes on X (formerly Twitter) following Elon Musk’s acquisition, focusing on whether the loosening of moderation policies and the introduction of monetization incentives have influenced the prevalence of hate speech and politically charged discourse. Central to this analysis is the question of how shifts in content regulation on a major social media platform intersect with democratic values and the principle of open public discourse. Drawing from theoretical foundations like the “marketplace of ideas” and the implications of algorithmic curation, the study examines the tension between free expression and the potential manipulation of online dialogue. The case of Musk’s acquisition offers a unique lens through which to study the dynamics of digital speech regulation. After taking control of X, Musk implemented sweeping changes, including reinstating previously banned accounts, introducing a monetized verification system, and dismantling long-standing transparency initiatives. These actions, justified under the banner of “free speech,” have significantly altered both the structure and substance of discourse on the platform. This thesis specifically examines the interactions surrounding four previously banned accounts, Donald Trump, Marjorie Taylor Greene, the Babylon Bee, and James Lindsay, analyzing changes in sentiment and toxicity pre- and post-reinstatement. Using sentiment analysis with RoBERTa and toxicity scoring with Google Perspective API, the study assesses whether users engaging with these reinstated accounts exhibit shifts in negativity or toxicity following Musk’s policy changes. The results indicate that while some users displayed increased negativity or toxicity, the findings varied across user characteristics such as verification status, political orientation, and follower count. Ultimately, although some changes in toxicity and sentiment after Musk’s acquisition of Twitter were statistically significant, they were not meaningful in a real-world sense. Beyond the empirical findings, the thesis discusses the broader implications of Musk’s control over X, highlighting the risks posed by the consolidation of algorithmic control within a single individual. As X transitions from a space of open public discourse to one shaped by opaque moderation practices and monetization incentives, the potential for algorithmic favoritism and the erosion of genuine democratic debate becomes apparent. The thesis also includes an analysis of existing transparency initiatives, a review of prominent case law, and considers legal theory put forth by leading legal scholars. After contextualizing the issue within this broader framework, the thesis proposes an amendment to existing legislation aimed at enhancing transparency in platform governance, carefully designed to avoid First Amendment infringements and address compliance challenges observed in previous transparency initiatives. Major themes addressed in this thesis include content moderation, algorithmic curation, free speech, hate speech, monetization incentives, political polarization, and the implications of private control over public discourseen-USPrinceton University Senior Theses