Anthropic Launches Project Glasswing for AI-Driven Critical Software Defense
Original: Project Glasswing View original →
On April 7, 2026, Anthropic announced Project Glasswing, an initiative designed to secure critical software with early access to Claude Mythos Preview. Anthropic said the program starts with launch partners including Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks.
The company said it is also extending access to more than 40 additional organizations that build or maintain critical software infrastructure. Anthropic committed up to $100 million in usage credits and $4 million in donations to open-source security organizations. According to the announcement, Claude Mythos Preview has already identified thousands of zero-day vulnerabilities across critical infrastructure and is being offered as a gated research preview.
Why the launch stands out
Project Glasswing is notable because it packages frontier-model access, cybersecurity evaluation, and industry coordination into one program. Instead of treating vulnerability discovery as a narrow internal benchmark, Anthropic is putting its model into defensive workflows at cloud providers, software vendors, financial institutions, and open-source organizations. The company is also making the model available through the Claude API, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry.
The announcement also underlines a broader shift in AI security. Multiple launch partners describe a much shorter window between vulnerability discovery and exploitation, especially as AI systems become better at understanding and modifying code. In that context, Anthropic is positioning Glasswing as an effort to give defenders and maintainers access before similar capabilities diffuse more widely.
Anthropic also disclosed pricing for participants, saying Claude Mythos Preview will be available at $25 per million input tokens and $125 per million output tokens. That detail makes the program more concrete: this is not just a research statement, but an operational channel for security teams and open-source maintainers to test AI-augmented defense in live environments.
Related Articles
Anthropic published a coordinated vulnerability disclosure framework on March 6, 2026 for vulnerabilities discovered by Claude. The policy sets a default 90-day disclosure path, a compressed 7-day path for actively exploited critical bugs, and a 45-day buffer after patches before technical details are usually published.
Anthropic introduced Project Glasswing on X and detailed the initiative on April 7, 2026 as a coordinated effort to secure critical software with Claude Mythos Preview. The launch matters because it treats defensive AI deployment as an industry-scale infrastructure problem, not just a model demo.
On April 7, 2026, Anthropic said on X that it has partnered with AWS, Apple, Google, Microsoft, NVIDIA, and others on Project Glasswing. Anthropic says the initiative gives selected defenders access to Claude Mythos Preview to find and fix critical software vulnerabilities, backed by up to $100 million in usage credits and $4 million in donations.
Comments (0)
No comments yet. Be the first to comment!