Assurance on Cognaptus https://cognaptus.com/tags/assurance/ Recent content in Assurance on Cognaptus Hugo -- 0.145.0 en-us Thu, 30 Apr 2026 00:00:00 +0000 Catch Me If You Can, Agent: Benchmarking AI That Learns to Look Safe https://cognaptus.com/blog/2026-04-30-catch-me-if-you-can-agent-benchmarking-ai-that-learns-to-look-safe/ Thu, 30 Apr 2026 00:00:00 +0000 https://cognaptus.com/blog/2026-04-30-catch-me-if-you-can-agent-benchmarking-ai-that-learns-to-look-safe/ A practical reading of ESRRSim, a taxonomy-driven framework for testing whether agentic AI systems can deceive, game evaluations, or manipulate oversight. Zero Degrees, Still Feverish: Why Deterministic AI Needs a Thermometer https://cognaptus.com/blog/2026-04-29-zero-degrees-still-feverish-why-deterministic-ai-needs-a-thermometer/ Wed, 29 Apr 2026 00:00:00 +0000 https://cognaptus.com/blog/2026-04-29-zero-degrees-still-feverish-why-deterministic-ai-needs-a-thermometer/ A business-focused reading of background temperature: a practical metric for measuring hidden randomness in LLM inference stacks, even when temperature is set to zero.