---
schema_version: '1.0'
id: news-20260816-ad13fe
title: 'Findings from a pilot Anthropic–OpenAI alignment evaluation exercise: OpenAI
  Safety Tests'
url: https://openai.com/index/openai-anthropic-safety-evaluation/
url_hash: ad13fe7d3321b824a6180da3456d9995ee9780de8135ee1147c7738a062f8990
canonical_url: https://openai.com/index/openai-anthropic-safety-evaluation
source: openai.com
category: news/tech
category_raw: technology
region: null
tags:
- technology
- artificial_intelligence
- machine_learning
- natural_language_processing
- ai_safety
- research
lang: null
published_at: null
fetched_at: '2026-08-16T13:00:52.114919Z'
updated_at: '2026-08-20T20:11:49.471933Z'
status: published
content_hash: null
content_changed_at: null
license_note: link-only
summary: null
summary_source: null
summary_en: null
summary_omitted_reason: license_restricted
redistribution: false
keywords_en:
- ai_safety
- alignment
- Anthropic
- artificial_intelligence
- Copilot AI
- evaluation
- exercise
- Findings
- from
- general news
- machine_learning
- natural_language_processing
- OpenAI
- pilot
- research
- Research Paper
- Safety
- Technology
- Tests
entities_en:
- name: Anthropic
  type: organization
- name: Copilot AI
  type: artifact
- name: Research Paper
  type: artifact
- name: Technology
  type: UNKNOWN
topics_en:
- general news
numeric_facts: []
entities:
- name: Technology
  type: UNKNOWN
- name: Research Paper
  type: artifact
- name: Copilot AI
  type: artifact
- name: Anthropic
  type: organization
key_facts: []
related: []
related_auto:
- name: AI技術
  type: concept
  weight: 1.0
- name: Microsoft Mico
  type: organization
  weight: 1.0
- name: turf_war
  type: event
  weight: 1.0
- name: 中国Z.ai
  type: organization
  weight: 1.0
- name: Opus 5
  type: concept
  weight: 1.0
---

# Findings from a pilot Anthropic–OpenAI alignment evaluation exercise: OpenAI Safety Tests

## TL;DR
（要約なし・原文はリンク先を参照 / No summary available; see the source link.）

## Key Points
- technology / artificial_intelligence / machine_learning / natural_language_processing / ai_safety / research

## Details
(本文なし。リンク先参照)

## Source
元記事: [Findings from a pilot Anthropic–OpenAI alignment evaluation exercise: OpenAI Safety Tests](https://openai.com/index/openai-anthropic-safety-evaluation/)