Unreleased Claude model finds attack chains in Chrome and Safari
Anthropic says an unreleased version of its Claude AI model found attack chains in Chrome and Safari, the web browsers used by millions. The finding suggests AI can now spot serious security weaknesses in everyday software, so browser makers and users may need to pay close attention.
Quick glossaryAnthropic, attack chains
- Anthropic
- Anthropic is an AI company based in San Francisco, best known for its Claude chatbot and AI models.
- attack chains
- A linked series of security flaws that, used one after another, lets a hacker take control of a device or steal data.
Burning Potato score: 70%