# How can you tell in system logs when an AI agent is trying to break rules, rather than a human hacker? — retained record

> Retained reddit record from reddit:cybersecurity.

Canonical: https://highsignal.app/data/records/51b661b32696dc06

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:cybersecurity)  published 2026-09-02 17:44 ingested 2026-09-03 02:38

[original source ↗](https://www.reddit.com/r/cybersecurity/comments/1w5hftj/how_can_you_tell_in_system_logs_when_an_ai_agent/)

## retained text



```
*Building log alerts, so trying to understand how autonomous behavior looks in practice*
```



document fetched 2026-09-03 00:24 · canonical https://reddit.com/r/cybersecurity/comments/1w5hftj/how_can_you_tell_in_system_logs_when_an_ai_agent
