Apple Joins Anthropic's Project Glasswing to Uncover Thousands of Software Flaws
Apple has joined Anthropic’s Project Glasswing alongside Amazon, Microsoft and Nvidia, using Claude Mythos Preview AI to identify thousands of software vulnerabilities across operating systems and browsers. The partnership aims to develop safeguards against AI-driven cyberattacks by integrating Mythos’s hacking-detection capabilities into Apple’s security protocols.
1. Project Glasswing Initiative
Anthropic launched Project Glasswing after its Claude Mythos Preview model unexpectedly excelled at identifying and chaining software vulnerabilities, scanning popular operating systems and web browsers to discover thousands of exploitable flaws. The initiative pools AI advancements to proactively test security defenses before advanced hacking models become widely available.
2. Apple's Role in AI-driven Cybersecurity
Apple joins Project Glasswing alongside Amazon, Microsoft and Nvidia to integrate Mythos’s hacking-detection outputs into its development pipelines. This collaboration is set to strengthen pre-release software testing and accelerate patch deployment, enhancing security posture across Apple's operating systems and applications.