---
format: "aidr-story-markdown/v1"
id: "0d4c9325e15ebfd9daad3e6443fc9f387f4a5a352316844e21c41c744051fef9"
canonical_url: "https://aidr.today/0d4c9325?lang=en"
title: "OpenAI: We monitor internal coding agents for misalignment"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-09-06T16:37:55.000Z"
category: "Agents"
topics: ["openai","agent","safety","agent-observability","coding"]
source_urls: ["https://openai.com/index/how-we-monitor-internal-coding-agents-misalignment/","https://news.ycombinator.com/item?id=49588214"]
summary: "How OpenAI uses chain-of-thought monitoring to study misalignment in internal coding agents—analyzing real-world deployments to detect risks and strengthen AI safety safeguards."
---

# OpenAI: We monitor internal coding agents for misalignment

> [Open the canonical story](<https://aidr.today/0d4c9325?lang=en>)

**Published:** 2026-09-06T16:37:55.000Z
**Category:** Agents
**Topics:** openai, agent, safety, agent\-observability, coding

## Summary

How OpenAI uses chain\-of\-thought monitoring to study misalignment in internal coding agents—analyzing real\-world deployments to detect risks and strengthen AI safety safeguards\.

## Sources

- [Story source](<https://openai.com/index/how-we-monitor-internal-coding-agents-misalignment/>)
- [Discussion](<https://news.ycombinator.com/item?id=49588214>)