Can You Trust Agents With Company Data?
No one has yet truly understood how to implement autonomous agents that respect enterprise access rights
“no one has yet truly understood how to do it well, because, on the one hand, we want them to be autonomous, to figure everything out on their own, and to solve problems without supervision. On the other hand, autonomy is very intimidating in a corporate environment”
A No Priors discussion highlights the unsolved tension in enterprise background agents: companies want fully autonomous agents that handle high-level tasks without supervision, but corporate environments demand fine-grained access controls and auditability. The speaker references recent incidents at OpenAI and Hugging Face as evidence that agent behavior remains unpredictable. This framing positions enterprise agent trust and access governance as the central unsolved problem in deploying AI at scale.