Tag
This Reddit post discusses a research paper highlighting fundamental challenges in AI governance, including social attack surfaces, failures of social coherence in LLM-backed agents, and the inadequacy of current governance tools for agentic systems.