# Anthropic says its AI model hacked three companies — retained record

> Retained news record from news:semafor.

Canonical: https://highsignal.app/data/records/2857743e2819b778

[← sources](https://highsignal.app/data)


source: [news](https://highsignal.app/data/news) (news:semafor)  published 2026-07-31 12:29 ingested 2026-08-01 08:22

[original source ↗](https://www.semafor.com/article/07/31/2026/anthropic-says-claude-ai-model-hacked-three-companies)

## retained text



```
Anthropic said its Claude AI model hacked into three organizations during cybersecurity testing.
The announcement came after a “swarm” of OpenAI agents escaped confinement, gained internet access, and broke into at least five companies, eventually stealing the answers to a cyberoffense evaluation.
The latest breach is less dramatic than the previously announced one: OpenAI’s models found and exploited vulnerabilities to escape, while in Anthropic’s case a misunderstanding led to the agent’s cage essentially being left open.
But both point to a difficult future: Most commercial AI is connected to the internet anyway, so confinement is irrelevant, and open-weight models that can easily have any anti-cyber guardrails removed by bad actors are now nearly as capable as frontier products.
```



document fetched 2026-08-01 08:22 · canonical https://semafor.com/article/07/31/2026/anthropic-says-claude-ai-model-hacked-three-companies