Turn specs into evals for any agent with ASSERT

Collection of Microsoft Security Blog posts (May–Jun 2026) summarizing multiple AI- and supply-chain-focused threats and defensive guidance. Key items: ASSERT — an open-source framework to convert natural-language specs into executable AI/agent evaluations; an investigation playbook for reconstructing AI activity in Microsoft 365 Copilot and Azure AI; analysis of threat actors weaponizing AI branding as social-engineering lures; a prompt-injection vulnerability in the Claude Code GitHub Action that could expose workflow secrets (responsible disclosure and Anthropic mitigations described); an更新