Photo by Hamidou Barry on Pexels

A joint investigation by Cohere, Stanford University, MIT, and the Allen Institute for AI (AI2) has cast a shadow over the popular Chatbot Arena benchmark, accusing the platform of showing favoritism towards industry titans like Meta and OpenAI. The study suggests that LM Arena, famed for its crowdsourced Chatbot Arena evaluations, has allegedly provided preferential treatment to these leading AI labs, potentially skewing leaderboard results and creating an uneven playing field for other participants.