In a significant move to combat rapidly evolving cyber threats, Amazon Web Services (AWS) has partnered with Anthropic to introduce a new generation of artificial intelligence tailored for cybersecurity. As threat landscapes grow increasingly complex, this collaboration aims to deliver smarter, faster, and more scalable security solutions for critical digital infrastructure.

As part of Anthropic’s newly announced Project Glasswing, the companies unveiled a specialized AI model called Claude Mythos Preview. Notably, this model has entered a gated release phase, specifically designed to protect highly sensitive and widely used software systems.

AWS has already established a strong foundation in AI-driven security. For instance, its internal machine learning-powered log analysis has dramatically reduced security review times—from an average of six hours to just seven minutes. Moreover, by processing over 400 trillion network flows daily, AWS managed to block more than 300 million malicious S3 encryption attempts in 2025 alone. Building on this success, the company is now extending these advanced capabilities to enterprise customers.

Claude Mythos Enters Preview

Claude Mythos Preview marks a major advancement in AI reasoning, particularly for cybersecurity and complex coding environments. It is specifically engineered to identify and remediate vulnerabilities at scale, thereby minimizing the need for manual intervention by security engineers.

However, given the inherent risks associated with AI models capable of generating exploit code, both AWS and Anthropic are proceeding cautiously. As a result, Claude Mythos Preview is currently accessible only through a gated research preview on Amazon Bedrock.

Access remains limited to a curated group of organizations, including internet-critical enterprises and leading open-source maintainers whose software serves millions globally. This selective approach ensures that the technology is tested responsibly while safeguarding broader ecosystems.

Furthermore, organizations using the model benefit from robust enterprise-grade protections. These include customer-managed data encryption for enhanced privacy, strict Virtual Private Cloud (VPC) isolation to secure workloads, and Automated Reasoning safeguards that deliver up to 99% accuracy against factual hallucinations.

Autonomous Penetration Testing Takes Center Stage

In addition to launching Claude Mythos, AWS has also made its AWS Security Agent generally available. This innovation transforms penetration testing from a periodic, manual process into a continuous, on-demand capability.

Operating around the clock, the Security Agent leverages specialized AI to independently detect and validate vulnerabilities across multi-cloud environments, including AWS, Azure, GCP, and on-premises systems. Unlike traditional scanners that often produce unverified alerts, this tool actively attempts to exploit vulnerabilities using targeted attack chains.

Once a genuine threat is confirmed, the system provides detailed, actionable insights. These include standardized CVSS risk scores, application-specific severity ratings, step-by-step reproduction guidance, and practical remediation strategies.

As cybercriminals, including nation-state actors and ransomware groups, increasingly adopt AI to scale their attacks, defensive strategies must evolve accordingly. Therefore, by combining Anthropic’s advanced AI capabilities with autonomous cloud security tools, AWS aims to empower organizations to proactively defend against emerging threats—before they even materialize.

Recommended Cyber Technology News:

To participate in our interviews, please write to our CyberTech Media Room at info@intentamplify.com  



🔒 Login or Register to continue reading