Anthropic unveils Project Glasswing to boost cybersecurity with AI: Details
Anthropic has announced Project Glasswing, a new cybersecurity initiative bringing together Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA and Palo Alto Networks to secure critical software systems using advanced AI. The company said the effort is driven by capabilities observed in its unreleased frontier model, Claude Mythos Preview, which can identify software vulnerabilities and help security teams fix them before they are misused by attackers.
What will Claude Mythos do
According to Anthropic, Claude Mythos Preview is capable of scanning code and systems to detect previously unknown security flaws, including critical vulnerabilities that may have gone unnoticed for years. The model can also simulate how these flaws could be exploited, allowing companies to better understand the risks and patch them more effectively.
Anthropic said the model has already identified thousands of high-severity vulnerabilities across major operating systems, web browsers and widely used software. These findings are being responsibly disclosed so that developers can fix them before any real-world misuse.
Regarding its availability, the company said, “We do not plan to make Claude Mythos Preview generally available, but our eventual goal is to enable our users to safely deploy Mythos-class models at scale—for cybersecurity purposes, but also for the myriad other benefits that such highly capable models will bring. To do so, we need to make progress in developing cybersecurity (and other) safeguards that detect and block the model’s most dangerous outputs. We plan to launch new safeguards with an upcoming Claude Opus model, allowing us to improve and refine them with a model that does not pose the same level of risk as Mythos Preview.”
What this means for users
For everyday users, the announcement points to a future where cybersecurity could improve significantly, even as threats grow. Anthropic noted that many digital services people rely on—from banking to healthcare—run on software that may contain hidden flaws.
With AI tools like Mythos Preview, companies can detect and fix these issues faster, potentially reducing the risk of data breaches, service disruptions and cyberattacks. The initiative also extends to open-source software, which powers many consumer apps and platforms.
Industry collaboration and next steps
Project Glasswing includes participation from the listed companies along with over 40 additional organisations working on critical infrastructure. Anthropic said it is committing up to $100 million in usage credits for the initiative, along with funding support for open-source security efforts.
The company emphasised that collaboration will be key as AI-driven threats evolve. It plans to share learnings from the project, including vulnerabilities fixed and best practices, to help improve cybersecurity standards across the industry.
Anthropic added that while advanced AI models introduce new risks, they can also play a crucial role in strengthening digital security when used responsibly.
