TikTok Algorithm Favored Conservative Content in 2024 POTUS Race, Study Suggests

A study conducted by researchers at New York University Abu Dhabi indicates that the TikTok recommendation algorithm favored politically conservative content during the 2024 U.S. Presidential election. The findings suggest that the platform pushed conservative-leaning material to users regardless of their individual political preferences or previous interactions.

The research was led by Talal Rahwan and Yasir Zaki, who sought to determine whether the platform’s “For You” page acted as a neutral mirror of user interest or if it actively steered users toward a specific ideological direction. The study focused on the period leading up to the November 2024 election to analyze how political information was disseminated to the American electorate.

To isolate the influence of the algorithm from human behavior, the researchers deployed more than 300 “dummy” bot accounts. These accounts were designed to mimic real-life users by simulating different political personas. Some bots were programmed to interact with liberal content, others with conservative content, and some remained politically neutral.

By automating the interactions—such as liking, sharing, and viewing specific types of videos—the team could observe how the algorithm responded to different inputs. The goal was to see if a bot that expressed a preference for liberal viewpoints would continue to receive a balanced feed or if the system would eventually introduce and prioritize content from the opposite end of the political spectrum.

Findings on Algorithmic Bias

The data collected by Rahwan and Zaki revealed a consistent trend toward conservative content. Despite the varied initial preferences of the bot accounts, the “For You” page began to skew more conservative over time across the majority of the simulated personas.

The researchers found that the algorithm did not simply reinforce the existing biases of the users—a phenomenon commonly known as the “filter bubble”—but instead appeared to promote conservative narratives more aggressively than liberal ones. This suggests a systemic bias within the recommendation engine that operated independently of the user’s demonstrated interests.

This outcome contradicts the common assumption that social media algorithms only create echo chambers by feeding users more of what they already like. In this instance, the NYU Abu Dhabi study suggests that the algorithm functioned as an active promoter of a specific political leaning, potentially exposing users to conservative viewpoints more frequently than they would have sought out on their own.

Implications for the 2024 Election

The timing of this algorithmic behavior is significant given TikTok’s role as a primary news source for younger voters during the 2024 presidential race. Because the “For You” page is the central feature of the user experience, any systemic skew in content delivery can influence the perception of political candidates and the prominence of specific policy issues.

The study highlights a potential vulnerability in how political information is curated on short-form video platforms. When an algorithm prioritizes one ideological direction over another, it can shift the “Overton Window”—the range of policies acceptable to the mainstream population—by making certain viewpoints appear more dominant or popular than they are in the general population.

the researchers noted that the lack of transparency regarding TikTok’s proprietary algorithms makes it difficult to determine if this skew was an accidental byproduct of engagement metrics or a deliberate design choice. Engagement-based algorithms often prioritize content that triggers strong emotional responses, and the study suggests that conservative political content may have triggered the specific engagement signals that the TikTok algorithm is programmed to reward.

Broader Context of Platform Influence

The findings from NYU Abu Dhabi add to a broader academic and regulatory debate regarding the influence of foreign-owned platforms on domestic political processes. TikTok has faced ongoing scrutiny from U.S. Lawmakers over concerns that its parent company, ByteDance, could influence the algorithm to serve specific political agendas.

While previous concerns often focused on the potential for the platform to suppress certain topics or promote foreign interests, this study points toward a specific domestic political bias during a critical election cycle. The result indicates that the platform’s influence may be more subtle than direct censorship, operating instead through the weighted promotion of specific ideological content.

The research underscores the difficulty of auditing black-box algorithms. Because TikTok does not provide external researchers with full access to its recommendation code, teams like those led by Rahwan and Zaki must rely on “black-box testing”—observing inputs and outputs via bot accounts—to infer how the system operates.

As of May 7, 2026, the study serves as a primary piece of evidence for those calling for greater algorithmic transparency and the implementation of third-party audits for social media platforms that reach millions of voters.