Beyond Bias: AI as a Proxy Advisor

After documenting a trend towards increasingly subjective proxy advisor voting guidelines, we evaluate the use of artificial intelligence as an unbiased proxy advisor for shareholder proposals. Using ISS guidelines, our AI model produces voting recommendations that match ISS in 79% of proposals and better predicts shareholder support than ISS recommendations alone. Disagreements between AI and ISS are more likely when firms disclose hiring a third-party governance consultant, suggesting these consultants-often the proxy advisor itself-may influence recommendations. These findings offer insight into proxy advisor conflicts of interest and demonstrate AI's potential to improve transparency and objectivity in voting decisions.