Anthropic has rolled out a preview of its new AI model, Mythos, as part of a broader cybersecurity initiative called Project Glasswing, bringing together major tech and infrastructure players including Nvidia, Google, Amazon Web Services, Apple, and Microsoft. Instead of a public release, the model is being shared in a controlled preview with select partners to test its ability to strengthen software security at scale.
This latest model is designed to help detect hidden security flaws in large software systems and cloud infrastructure before attackers can find them. Early use is focused on scanning codebases and identifying complex vulnerabilities that traditional tools may miss. The goal of the project is to speed up vulnerability detection and improve software security across widely used platforms.
Beyond basic bug detection, Mythos is positioned as a more advanced reasoning system for security analysis. Rather than simply flagging known patterns and common coding errors, it is intended to examine how different parts of a system interact, identify subtle logic flaws, and trace how small weaknesses could potentially combine into more serious exploit chains.
Notably, Project Glasswing is structured around the idea that cybersecurity is increasingly a large-scale coordination problem. Modern digital infrastructure is updated continuously, and vulnerabilities can appear not only in proprietary software but also in widely used open-source dependencies. And by deploying Mythos in a controlled setting with major cloud providers and technology companies, the initiative aims to create a shared layer of defense where security issues can be identified earlier in the development cycle. A key focus of early testing is large-scale code review and infrastructure scanning. In practice, this means Mythos is being applied to complex enterprise systems that would normally require extensive manual security auditing.
Another important aspect of the initiative is the focus on cloud security. As more organizations rely on distributed computing platforms, misconfigurations and subtle permission issues have become a major source of security breaches. Within Project Glasswing, Mythos is being evaluated for its ability to analyze cloud architecture setups, identify insecure configurations, and suggest corrective changes before those systems are deployed.
“Mythos Preview has already found thousands of high-severity vulnerabilities, including some in every major operating system and web browser. We have also extended access to a group of over 40 additional organizations that build or maintain critical software infrastructure so they can use the model to scan and secure both first-party and open-source systems. Anthropic is committing up to $100M in usage credits for Mythos Preview across these efforts, as well as $4M in direct donations to open-source security organizations,” the Dario Amodei-led AI firm noted.
The Tech Portal is published by Blue Box Media Private Limited. Our investors have no influence over our reporting. Read our full Ownership and Funding Disclosure →
Ashutosh is a Senior Writer at The Tech Portal, largely reporting on new tech, and intersection of technology and business. Ashutosh’s career spans across nearly a decade of technology writing across multiple platforms and languages.
