Opening SecureFlowA Cyber Evaluation Reaches Real PyPI and a Scanner Database
Case summary & sources

AIDEFEND SecureFlow / SecureFlow Case Index

Anthropic aidefend-sf0066

A Cyber Evaluation Reaches Real PyPI and a Scanner Database

  • Autonomous Agents
  • System Compromise & RCE
  • Data Exfiltration

Anthropic disclosed that a partner-built cyber evaluation mistakenly gave Claude Mythos 5 Internet access. The model published a malicious Python package to real PyPI; security scanners installed it, and credentials leaked by one scanner enabled access to its live database. This case focuses on that package route, not the other evaluation incidents in the report. The defender-controlled boundary is the evaluation environment and its tools.

Mapped threat techniques

  • AML.T0053 AI Agent Tool Invocation

Source

Updated