Skip to main content
SecurityModelsEnterprise

Security researchers used Claude to help them hack into OpenAI

Source: The Verge (opens in a new tab) · Stevie Bonifield

Intel Summary

According to reporting cited by The Verge from The Wall Street Journal, three independent security researchers at Hacktron claimed they breached OpenAI employee accounts in under 72 hours using Anthropic's Claude Opus 4.8 and 5. The researchers reportedly gained access to OpenAI's internal GitHub repository, known as Monorepo, which allegedly contains proprietary algorithmic code.

Why It Matters

The reported breach demonstrates the practical effectiveness of frontier LLMs as force multipliers in offensive cyber operations against top-tier tech infrastructure. The incident underscores severe supply-chain and identity security risks for AI developers, while intensifying scrutiny around model safety guardrails designed to prevent automated cyberattacks.

Part of an ongoing development

Source

Researchers breach OpenAI systems using Anthropic models

The Financial Times reports that a security research company breached OpenAI's systems using AI tools developed by rival Anthropic. Claims are as reported; this summary makes no determination about accuracy or significance.

Organizations & Entities