It seems Apple has decided to recruit a new “genius” to its security team, but this time, it is not human. Anthropic recently announced a new initiative called “Project Glasswing,” aimed at helping tech giants discover and fix security vulnerabilities before malicious actors find them. The idea is both simple and daunting: using a new AI model called Mythos Preview that outperforms the most skilled human programmers in hunting for software weaknesses.
Project Glasswing: The vulnerability hunter that never sleeps
Project Glasswing is not just a technical experiment; it is an urgent attempt to put AI capabilities at the service of defense rather than attack. According to Anthropic, the Mythos Preview model has already succeeded in discovering thousands of “zero-day” vulnerabilities (vulnerabilities unknown to companies), which existed in all major operating systems and web browsers we use daily.
AI has now reached a level of professionalism in writing and reading code that surpasses human capabilities. With the rapid pace of development, it has become essential for major companies like Apple to possess these defensive tools before extortionists and hackers—who care little for security or privacy standards—get their hands on them. It is like an arms race, but the weapons here are lines of code and programming logic.
Apple and the great security alliance
Apple is not alone in this exclusive club. Access to the Mythos Preview model will not be available to the public (so we don’t find our vulnerabilities being sold on the market), but will be limited to a select group of partners. The list includes, alongside Apple, companies like Google, Microsoft, Amazon, and Nvidia. Anthropic is quite generous here, having allocated up to $100 million in usage credits for partners; after that, Apple will have to pay the bills to keep this “smart inspector” running.
Apple will use this technology to enhance the security of the Safari browser and its various systems, from the iPhone system to macOS, watchOS, and even the visionOS for its headset. The goal is for the AI to read millions of lines of code in seconds, telling engineers in Cupertino: “Sorry, you have a vulnerability here that might allow hackers to get in; it’s better to close it now before it’s too late.” And since rumors are talking about a new homeOS, it is good that this system will begin its life under the watch of strict AI security.
Source:
Leave a Reply