Exploring and Mitigating the Impact of Increasingly Algorithmic Social Media Recommender Systems on U.S. Political Polarization

This project provides an overview of the current landscape of political polarization in the U.S. and current social media recommender systems, then proposes two interventions.

MIT D-Lab Class

Entrepreneurship for the Idealist


Elaine Sun Wellesley '23: Economics & Women's and Gender Studies student interested in economic and social inequality. Also avid crossword puzzler.



Problem framing

Social media recommender systems are growing increasingly algorithm-driven, and with the potential implementation of advanced AI systems into these algorithms, the media that users see and interact with could change drastically. In addition, Democrats and Republicans in the U.S. are now farther apart ideologically compared to any time in the last 50 years. The combination of increasingly algorithmic recommender systems and the implementation of AI may have serious implications on the content shown to social media users and play a role in exacerbating political polarization in the United States.

Proposed solutions

  1. Social media ranking system based on a polarization factor and/or level of misinformation
  2. Media consumption habits analyzer


Manish Bhardwaj, Entrepreneurship for the Idealist Instructor