MIT research examines AI's effects on learning and election information
MIT-backed projects are investigating how generative AI affects learning, cognition and access to political information. One ongoing study found that voters researching Alaska's midterm elections with Claude may have received incomplete candidate information.
AI Hubs United Editorial ·
MIT researchers are studying how generative AI changes learning and access to political information, with support from the MIT Human Insight Collaborative (MITHIC). According to MIT AI News, roughly one-third of the initiative's 67 supported projects focus on AI, society and humanity. Since its launch in 2024, MITHIC has provided $5.5 million across its full portfolio, spanning 31 MIT units.
One ongoing project involves political scientists Adam Berinsky and Charles Stewart III and MIT Sloan Assistant Professor Chara Podimata. It examines large language models as sources of democratic knowledge in the context of the 2026 U.S. midterm elections. According to the report, the project demonstrated that Alaska voters using Anthropic's Claude for election research may have received incomplete information about candidates and their positions. The findings raise practical concerns about relying on AI tools for political research.
Separately, Eric Klopfer and Pawan Sinha are measuring AI's short- and long-term effects on cognition. The work comes alongside an MIT education report documenting less frequent student attendance at office hours and less time spent on problem sets and with peers, prompting concerns about AI's effects on learning and social interaction.
AI Hubs United original short summary of the source linked above. It is not the original article; read the full story at the source.
