Expired50% confidenceReleaseExact time

OpenAI disclosed an internal adversarial model called GPT-Red, a red-teaming security tool designed specifically to attack AI agents.

1

Sources

50%

Confidence

Medium-term (~90 days)

Relevance

7/16/2026

First Seen

This claim has expired. It is kept for the record and no longer describes the current state.

Valid until: 10/14/2026(expired)

Sources

GPT-Red: OpenAI's Internal Red-Team Security Tool That Uses AI to Attack Its Own Models

Related Claims

UnverifiedOpenAI completely isolates GPT-Red from deployed models to prevent attack capabilities from leaking or being made available to potential adversaries.81% similar VerifiedOpenAI released GPT-5.5 Cyber, an AI model specifically targeting critical infrastructure defenders.75% similar UnverifiedOpenAI的InstructGPT、Anthropic的Constitutional AI以及Meta的Llama系列都采用了类似的安全对齐流程72% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/533953

API

curl https://kongchang.com/api/v1/knowledge/claims/533953

MCP

get_claim(id=533953)