# Pain Points in deploying Ai agents. — retained record

> Retained reddit record from reddit:AI_Agents.

Canonical: https://highsignal.app/data/records/e3698227fe870029

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:AI_Agents)  published 2026-11-02 06:52 ingested 2026-11-03 02:37

[original source ↗](https://www.reddit.com/r/AI_Agents/comments/1xlqar6/pain_points_in_deploying_ai_agents/)

## retained text



```
Ai agents have helped to bridge the gaps especially for solo founders, but when we talk about the security and authenticity there are pain points that are evolving. The challenge of unexpected responses and fabrications that keeps repeating even after setting proper guardrails has been considered as the biggest hindrance. 

This is vital when ai agents need seamless communications especially for enterprises and health care where marketing, accounts and managing patient data needs to be in strict sand boxes. The infra on which agents act need to be reliable and compliant to industry standards, with evolving legal and business landscape a failure to implement minute changes may result in substantial economic and reputation risks.

Human in the loop would be key when the the role of the executing agent stops to the level where human approval decides the future outcome. Agentic system with right amount of constraint is a challenging subject in itself, a too tight guardrails affects with excessive human filtering for automated tasks and too low may let the agents loose. A logical solution would be handling over the decision making.
```



document fetched 2026-11-03 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1xlqar6/pain_points_in_deploying_ai_agents
