# Only 22% of surveyed teams treat AI agents as independent identities. How are you handling yours? — retained record

> Retained reddit record from reddit:AI_Agents.

Canonical: https://highsignal.app/data/records/197d15728a360f4a

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:AI_Agents)  published 2026-10-13 12:40 ingested 2026-10-14 02:36

[original source ↗](https://www.reddit.com/r/AI_Agents/comments/1x4vjxt/only_22_of_surveyed_teams_treat_ai_agents_as/)

## retained text



```
As AI agents start working across different tools and systems, there's another question to figure out: should an agent have its own identity, or should it operate using the same credentials as the person who set it up?  
A Cloud Security Alliance study cited in this article found that only about 22% of surveyed teams treat AI agents as independent identities.  
For those building or deploying agents, how are you approaching this?

* Does each agent have its own identity and permissions? 
* Are agents using shared API keys or credentials inherited from a user? 
* Does your approach change depending on what the agent needs to do? 

If you're running multiple agents, would you give each one its own identity from day one, or only introduce that as your setup gets more complex? Why?
```



document fetched 2026-10-14 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1x4vjxt/only_22_of_surveyed_teams_treat_ai_agents_as
