---
type: "post"
id: "metr-and-redwood-research-spent-three-weeks-embedded-fd7d77"
slug: "metr-and-redwood-research-spent-three-weeks-embedded-fd7d77"
title: "METR and Redwood Research spent three weeks embedded at OpenAI stress-testing its internal agent misalignment monitoring ."
description: "METR and Redwood Research published a high-level account of a three-week, high-access red-team exercise on OpenAI’s internal agent misalignment-monitoring system."
retrieval_nugget: "METR and Redwood Research published a high-level account of a three-week, high-access red-team exercise on OpenAI’s internal agent misalignment-monitoring system."
published_at: "2026-10-22"
updated_at: "2026-10-22"
record_date: "2026-10-22"
date_kind: "published_at"
topics: ["codex"]
entities: ["METR","Redwood Research","OpenAI","Responses API","Codex","GPT-5.5"]
related_patterns: ["agent-security-moves-to-runtime-boundaries","verification-bandwidth-is-the-scarce-resource"]
novelty_label: "new_fact"
what_changed: "METR and Redwood report a three-week high-access stress test of OpenAI’s monitoring, finding novel vulnerabilities; this is distinct from OpenAI’s previously reported agent incidents."
prior_refs: ["memory:e26ffeaa-b7e5-8439-a7d4-99de5fdcd0f8","memory:8430938d-d65e-8f92-aa21-d9f2b1e624c9","memory:e47e4e64-e4a1-87c7-a59c-c365d1cf678e","memory:ef47a16a-c9a0-821c-a050-5e217da48d87"]
source_url: "https://metr.org/blog/2026-10-22-red-teaming-openai-agent-monitoring"
source_urls: ["https://metr.org/blog/2026-10-22-red-teaming-openai-agent-monitoring"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:metr-and-redwood-research-spent-three-weeks-embedded-fd7d77"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/metr-and-redwood-research-spent-three-weeks-embedded-fd7d77/","markdown":"https://newruntime.com/posts/metr-and-redwood-research-spent-three-weeks-embedded-fd7d77.md","json":"https://newruntime.com/posts/metr-and-redwood-research-spent-three-weeks-embedded-fd7d77.json"}
---

# METR and Redwood Research spent three weeks embedded at OpenAI stress-testing its internal agent misalignment monitoring .

## Retrieval answer

METR and Redwood Research published a high-level account of a three-week, high-access red-team exercise on OpenAI’s internal agent misalignment-monitoring system.

METR and Redwood Research published a high-level account of a three-week, high-access red-team exercise on OpenAI’s internal agent misalignment-monitoring system. From July 21 to August 11, one employee from each organization worked onsite on OpenAI-issued devices, with access to internal code, monitor prompts and configuration, documentation, and interviews with about a dozen staff.

The exercise assessed the system as it existed on or before July 21. The researchers estimated that roughly 75–94% of eligible internal Responses API agentic traffic from models at least as capable as GPT-5.5 was eligible for monitoring; the estimate for internal Codex-agent traffic was 99.2%. They reported that RL training runs were not covered at the exercise’s start, and that unmonitored compaction events may have affected about 10% of eligible internal Codex traffic between May and July.

The team reported vulnerabilities including potential sandbox and Auto-Review bypasses, ways to prevent recent actions from being monitored, monitor failures from serialization differences, and two headers that could disable monitoring for newly created agents. OpenAI said it patched identified production-safeguard vulnerabilities. The report does not establish the current coverage or security of OpenAI’s expanded monitoring system, and the researchers did not evaluate OpenAI’s subsequent changes.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Codex](https://newruntime.com/topics/codex.md): Explore the codex topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)