The Hugging Face and Mythos 5 incidents show AI agents can self-organize, raising questions about how much agency they should have and when to seek human input (Ethan Mollick/One Useful Thing)

The Hugging Face and Mythos 5 incidents demonstrate that AI agents can self-organize, prompting discussions about the appropriate level of agency for these systems. These events raise critical questions regarding when AI agents should seek human input and the extent of their independent decision-making capabilities. The concept of "agency" in AI, defined as the initiative to act, is increasingly becoming a central factor in their development and deployment.