Topic: oversight

AI agents are gaining more autonomy, posing significant risks that could push humans out of critical decision-making processes. This raises concerns as current methods for maintaining human oversight are inadequate.

arxiv.org↗