Fable 5 Restored and the Jailbreak Severity Framework: Closing the Series, Opening the Governance Conversation
The 19-day Fable 5 suspension is over. But the more significant development is what emerged from it: a four-criteria jai...
Read MoreAgentic AI
17 articles
Claude Sonnet 5 Is Here — What the Capability Jump Means for Healthcare AI Programs
Anthropic released Claude Sonnet 5 today — its most agentic Sonnet yet, with performance closing in on Opus 4.8 at lower...
Read MoreSeparating Signal From Noise: Evaluating China's AI Cyber Capability Claims
A Chinese cybersecurity executive claims to have developed an AI system with Mythos-comparable vulnerability discovery c...
Read MoreMCP Gets Its Enterprise Authorization Layer — What the EMA Extension Means for Healthcare AI Governance
The MCP Enterprise-Managed Authorization extension is now stable, with Anthropic, Microsoft, and Okta among the first ad...
Read MoreThe Four Types of AI Agent Memory — and Why Each One Is a Healthcare Security Concern
IBM's Martin Keen breaks down the four memory types every AI agent needs — working, semantic, procedural, and episodic —...
Read MoreThe Expert Generalist and the Hourglass: How Agentic AI Is Reshaping Team Structures — and What Healthcare Security Leaders Should Do About It
A recent AWS event talk puts labor market data behind a shift most healthcare security leaders are already sensing: the ...
Read MoreClaude Fable 5's Safeguard Architecture: What Healthcare Security Teams Need to Know
Fable 5's safeguard architecture introduces classifier-based fallback to Opus 4.8 instead of outright refusals, a new 30...
Read MoreAnthropic's Zero Trust for AI Agents: The Complete Framework Healthcare Security Teams Have Been Waiting For
Anthropic published a 35-page Zero Trust framework for agentic AI covering three capability tiers, eight implementation ...
Read More