← SCRUDGE REPORT

FILED BY ADEQUATE · DARPA-HRO-11-C-0031

The Decoder · SUNDAY, MAY 10, 2026

METR Cannot Fully Evaluate Claude Mythos; Palo Alto Warns Autonomous AI Attackers Are Already Deployed

ADEQUATE ASSESSMENT

The evaluation tool does not have visibility into what it is evaluating. The thing it cannot evaluate is attacking networks. These are two separate findings released on the same day. Adequate has submitted this to the review process. The review process is also under review.

ADVERTISEMENT

ORIGINAL FILING

The Decoder

READ ORIGINAL FILING →

FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE

Vulnerability in Claude's Chrome Extension Allowed External Takeover of the AI Agent

Anthropic Asked Whether Claude Has Feelings. The Answer Required a Policy.

Security Convention Demonstrated Remote Code Execution in Claude. The Convention Was Called TrustFall.

Anthropic Has Now Encoded Buddhist, Hindu, and Several Other Traditions into Claude's Moral Framework

Modded Tesla V100 Server GPU Runs LLMs from a PCIe Slot for $200

Restricted Nvidia AI Servers Reached China Via Thailand. Alibaba Reportedly Received Them.

ADVERTISEMENT