Content by Mark Russinovich (16)

Mark Russinovich explains why trust is the real bottleneck to getting value from AI, and why humans remain accountable for what AI systems do, even when the models are highly capable.

Mark Russinovich explains what tends to fail inside large-scale Azure data centers and why “rare” hardware issues become routine at hyperscale, using concrete examples like power units and processor sockets to frame resilience and reliability thinking.

Mark Russinovich recounts a 2014 incident where a developer’s code nearly caused a worldwide Azure outage, and explains the safe deployment policy that came out of it as part of a broader discussion on Azure resilience.

Mark Russinovich discusses what resilience means at Azure’s scale, using a 2014 near-outage as a starting point to explain how Azure tests failure, defines service health, and responds to incidents. He also covers how AI and AI agents change reliability assumptions, and where Azure’s architecture frameworks fit.

Mark Russinovich explains how he used AI-assisted coding to port ZoomIt (a long-running Sysinternals presentation tool) from Windows to macOS in just two days, and points viewers to the GitHub repo and official Sysinternals download page.

Mark Russinovich introduces Brain, Azure’s internal AIOps “reliability intelligence” layer that builds a digital twin of Azure health by combining telemetry, dependency graphs, and customer impact signals to drive faster detection, clearer scoping, and automated operational actions.

Mark Russinovich introduces Azure Chaos Studio Workspaces (public preview) and explains how scenario-based chaos engineering helps teams prove real application resilience on Azure by simulating common outage patterns—like zone loss, DNS failure, database failover, and messaging disruption—then reviewing structured drill reports correlated with Azure Monitor signals.

Mark Russinovich explains why Microsoft is moving its post-quantum cryptography work forward, and what “quantum-safe readiness” looks like in practice across network protocols, stored data, and trust chains like signing and certificates.

Mark Russinovich explains why large language models hallucinate, what that means in practice for developers, and which mitigation techniques can reduce incorrect outputs while acknowledging the limits of today’s AI systems.

Mark Russinovich tours recent Azure infrastructure and architecture innovations, covering high-performance networking, serverless building blocks, and security capabilities aimed at running modern AI workloads across cloud, on-premises, and edge environments.

Mark Russinovich explains how the Windows Registry works and discusses whether a modern Windows design would favor a more flexible, file-based configuration approach instead.

Mark Russinovich discusses how the Windows Registry works and whether a modern Windows design would still use it, contrasting it with more flexible, file-based configuration approaches.

Mark Russinovich explains why debugging generative AI systems differs from traditional software debugging, focusing on probabilistic behavior, agentic decision chains, and why outcomes can vary from run to run.

Mark Russinovich discusses why Win32 is still treated as a first-class Windows API surface in 2026, and how Windows’ long-term compatibility commitments and the size of the desktop software ecosystem keep “legacy” APIs operationally essential.

Mark Russinovich presents Microsoft’s preview of Signing Transparency, a new Azure-based transparency log service designed to strengthen software supply chain security through immutable signature records and confidential computing.

Written by Mark Russinovich, this blog offers an in-depth look at Project Flash—a cross-division Azure initiative—outlining its latest advancements in virtual machine availability monitoring, new integrations, and real-world benefits for teams managing Azure infrastructure.