AWS previews Claude Mythos

AWS has launched Claude Mythos in a gated Bedrock preview aimed at allowlisted organizations for advanced cybersecurity and code analysis, and introduced an Agent Registry for managing AI agents. The offering is positioned for restricted enterprise use rather than broad public access. (x.com)

Amazon Web Services has opened a gated research preview of Claude Mythos in Amazon Bedrock, limiting access to an allow-list of organizations instead of a broad customer rollout. (aws.amazon.com) Amazon said the model became available on April 7, 2026 as part of Project Glasswing, and its Bedrock documentation says access is “prioritized for defensive cybersecurity use cases.” (aws.amazon.com, docs.aws.amazon.com) In plain terms, Bedrock is Amazon’s managed marketplace for outside artificial intelligence models, and Mythos is being pitched for scanning software, analyzing code, and running longer multi-step tasks that act more like an automated teammate than a one-shot chatbot. (aws.amazon.com, docs.aws.amazon.com) Anthropic said Project Glasswing is aimed at “the organizations responsible for the infrastructure billions of people depend on,” with launch partners including Amazon Web Services, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA, and Palo Alto Networks. (anthropic.com) Anthropic also said it extended access to more than 40 additional organizations that build or maintain critical software infrastructure, and committed up to $100 million in usage credits plus $4 million in donations to open-source security groups. (anthropic.com) Amazon paired the model preview with a second release on April 9: AWS Agent Registry, a preview feature in Amazon Bedrock AgentCore that gives companies a private catalog for agents, tools, skills, Model Context Protocol servers, and other custom resources. (aws.amazon.com) That registry is meant to solve a practical enterprise problem: once companies start building many artificial intelligence agents, they need one place to discover them, track ownership, and apply governance controls before those agents spread across teams. (aws.amazon.com) Amazon’s security blog said organizations that were allow-listed for Mythos would be contacted directly by their Amazon Web Services account teams, underscoring that this is a restricted enterprise program rather than a public self-serve launch. (aws.amazon.com) Anthropic said participants can reach Mythos through its own application programming interface, Amazon Bedrock, Google Cloud Vertex AI, and Microsoft Foundry, with pricing listed at $25 per million input tokens and $125 per million output tokens after the initial program setup. (anthropic.com) The immediate result is a narrow release aimed at companies that defend critical software, while Amazon builds the controls to manage the agents those companies may deploy around it. (aws.amazon.com, aws.amazon.com)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.