Claude Platform on AWS: Understanding the Data Boundary for Healthcare AI
Anthropic's Claude Platform on AWS offers unified billing and AWS credentials, but routes data through Anthropic's infra...
Read MoreAI Security & Healthcare Insights
Practitioner-focused analysis at the intersection of AI security, AI governance, and healthcare cybersecurity
The Mythos 'Breach' That Wasn't: What Healthcare Learns About Vendor AI Risk
When Anthropic's 'too dangerous to release' Mythos AI leaked within hours of announcement, headlines blamed sophisticate...
Read MoreThe Malicious Middleman: How LLM Relays Became Healthcare's Newest Supply Chain Attack Vector
New research analyzed 428 LLM relay servers and found 9 actively injecting malicious code into AI tool calls. For health...
Read MoreSmall Models, Big Impact: Why Healthcare Doesn't Need Mythos to Find Vulnerabilities
One day after Anthropic announced Claude Mythos was too dangerous to release, security startup AISLE showed that $0.11/M...
Read MoreThe AI Gateway Everyone Uses Just Got Backdoored: LiteLLM and the Healthcare Supply Chain Risk
LiteLLM, the Python library with 95 million monthly downloads powering nearly every AI agent framework, was compromised ...
Read MoreUK Government Reality-Checks Claude Mythos: Why Healthcare's Cyber Basics Just Became Non-Negotiable
The UK AI Safety Institute's independent evaluation of Claude Mythos Preview reveals critical nuances missing from vendo...
Read MoreComment-and-Control: GitHub-Integrated AI Agents Vulnerable to Credential Theft
AI agents integrated with GitHub Actions can be hijacked via prompt injection to steal API keys and tokens. Anthropic, G...
Read MoreVS Code 1.115.0 Brings Parallel AI Agents to Healthcare Development—With Security Built In
Microsoft's VS Code 1.115.0 introduces parallel AI agent sessions with worktree isolation, permission controls, and audi...
Read MoreThe Underground AI Threat Healthcare Isn't Prepared For: GTG-1002 and What Comes Next
While healthcare debated AI governance, adversaries built autonomous attack systems. GTG-1002—the first documented AI-or...
Read More