In the rapidly evolving landscape of cybersecurity, groundbreaking news emerges from a collaboration between two tech giants. Anthropic, a leader in artificial intelligence, has announced the launch of Project Glasswing, a bold initiative designed to harness the power of its cutting-edge AI model, Mythos Preview, to discover and neutralize digital threats across operating systems and web browsers. In this context, Apple positions itself as a key partner, integrating this innovative technology to bolster the security of its ecosystems, including iOS, macOS, and Safari.
The Power of Mythos Preview in the Vulnerability Hunt
Mythos Preview has already demonstrated extraordinary capabilities, identifying thousands of zero-day vulnerabilities, some of which are present in major operating systems and browsers. Anthropic itself highlights how AI models have reached a level of coding proficiency that surpasses even the most skilled human experts in discovering and exploiting software weaknesses. Faced with this exponential progress, the company warns that such capabilities will proliferate rapidly, potentially beyond the control of actors committed to deploying them safely. Project Glasswing thus arises as an urgent attempt to employ these capabilities for defensive purposes.
Apple Invests in Proactive Defense
For Apple, adopting Mythos Preview represents a significant step towards proactive defense. The Cupertino giant will leverage the AI model to thoroughly scan for potential security flaws in Safari, its web browser, and across all its operating systems, including iOS, iPadOS, macOS, watchOS, tvOS, and visionOS. There are also rumors that Apple is developing an operating system named homeOS for a future smart home hub, an area that could greatly benefit from this new security technology. Anthropic's investment of up to $100 million in usage credits for Mythos Preview underscores the strategic importance of this partnership. Beyond this initial allocation, partners will incur costs for continued use of the AI model, highlighting the intrinsic value of the solution.
Implications for Global Digital Security
The collaboration between Anthropic and Apple goes beyond merely optimizing Apple's products. It marks a turning point in the approach to cybersecurity on a large scale. In an era where threats evolve at breakneck speed, employing advanced AI to identify vulnerabilities before they are exploited by malicious actors becomes crucial. This collaborative approach, which also involves other giants like Amazon Web Services, Google, and Microsoft, could usher in a new era of predictive security. While models like Anthropic's Claude 3 continue to set new standards in conversational AI, this initiative demonstrates how AI can also be applied to strengthen critical digital infrastructure. It is important to note that while AI is becoming a powerful tool for collaboration and innovation, as seen with Atlassian Confluence integrating Visual AI Tools, its application in security requires particular attention, especially concerning child protection, an issue on which OpenAI is actively working, and the regulation of social media for minors, as evidenced by recent discussions in countries like Australia Countries Worldwide Ban Social Media for Minors The Australian Case and Reasons Behind the Digital Prohibition.
Mythos Preview will not be available to the general public. Its distribution is reserved for selected partners, confirming the strategic and sensitive nature of this tool. The move by Anthropic and Apple not only promises to make systems more secure but also raises important questions about the ethics and future of AI development, a field that continues to surprise with innovations like foldable screens for mobile devices Apple Foldable Samsung Screen Deal for 3 Years and new interfaces for infotainment systems like WhatsApp Revolutionizing CarPlay.
Source: https://www.macrumors.com/2026/04/08/apple-anthropic-project-glasswing