Anthropic grants Apple, Amazon and other tech giants early access to its unreleased Mythos AI model to identify security vulnerabilities before wider release.

Anthropic launched Project Glasswing, partnering with Amazon, Apple, Microsoft, Cisco and other organizations to test the new Mythos AI model

The initiative aims to help identify potential cyberattacks and security risks that could emerge when the advanced AI system becomes more widely available

Selected tech firms receive early access to the unreleased model to conduct security testing and vulnerability assessments