IT a.m. 🌤️ Apr 20, 2026
Insurance Carriers Pull Back From Covering AI Outputs
Major insurance carriers are quietly excluding AI-related incidents from cybersecurity and general liability policies, according to CIO.com. Companies deploying AI for internal processes are finding that traditional cyber insurance won’t cover losses tied to AI-generated errors, hallucinations, or flawed decisions – forcing CIOs to rethink their AI risk strategies.
- According to CIO.com, insurers view AI outputs as unpredictable and difficult to underwrite, leading to broad exclusions in policy renewals. This mirrors how early cloud adoption faced similar coverage gaps before risk models matured.
- The coverage gap hits hardest for enterprises using AI in customer-facing or decision-critical workflows. Without insurance backstops, organizations bear full financial liability for AI failures, raising the stakes on governance and testing.
- For IT leaders, this signals a need to formalize AI risk frameworks, invest in output validation, and engage legal and risk teams before scaling deployments. Expect insurers to eventually offer AI-specific policies – but at premium pricing.
💻 IT News
Microsoft issued out-of-band updates to resolve issues affecting Windows Server systems after April 2026 security updates were installed.
🌩️ Cloud News
The European Commission awarded a €180 million ($212 million) sovereign cloud contract to four European providers for a six-year term. The tender supports the Commission’s efforts to enhance sovereignty and reinforce control.
AWS released DevOps Agent, a generative AI-powered assistant designed to help developers troubleshoot issues, analyze deployments, and automate operational tasks across AWS environments.
🧠 AI News
Sources say the US NSA is using Mythos Preview, and it’s widely used within the DoD, despite Anthropic’s supply chain risk designation.
OpenAI’s updated Agents SDK adds sandboxing and a model-native harness, giving enterprises a more controlled way to build and deploy long-horizon AI agents.
🛡️ Cybersecurity News
NIST will cease assigning severity scores to lower-priority vulnerabilities due to increased workload from rising submission volumes.
⚙️ DevOps News
Leapwork infuses agentic AI capabilities into its test automation platform to enable continuous validation across application testing workflows.
🏗️ Infrastructure News
SK hynix started mass production of the 192GB SOCAMM2, a next-generation LPDDR5X low-power DRAM module designed for Nvidia’s Vera Rubin.
📈 M&A News
Returning backers a16z and Thrive are expected to lead the funding round as Cursor’s enterprise growth surges.
🏢 Top IT Portfolio Rationalization Firms
According to The New Stack, Agentic ITOps Could End Costly Rules-Based Ops
According to The New Stack, rules-based IT operations cost businesses hundreds of billions annually through rigid automation that breaks under novel conditions. The shift to agentic ITOps – where AI agents reason, learn, and act autonomously – represents the next operational paradigm.
- The article outlines a maturity model moving from reactive (manual) to proactive (rules-based) to autonomous (agentic), urging leaders to assess where their teams currently sit before investing.
- According to the piece, preparation starts with high-quality observability data; agents are only as effective as the telemetry and context they can access.
- The author stresses that agentic does not mean unsupervised – human-in-the-loop governance, clear guardrails, and incremental trust-building are essential to safe adoption.
🔎 Dive Deeper
The agentic AI era is forcing a reset in enterprise architecture as agents that take action introduce requirements most enterprises are not engineered to handle. AI agents operating at machine scale require new architectural approaches beyond traditional data analysis.
Industry experts discuss cultural red flags threatening 2026 innovation goals for CIOs four months into the year. The article addresses how CIOs can identify and overcome organizational cultural barriers.
Developers are producing more code, but it’s more expensive and requires significantly more rewriting, reducing actual productivity.
Living off the Land attacks weaponize trusted tools like PowerShell, making them one of the most persistent threats facing enterprise security teams.
Study.com finds 9 in 10 employees use AI at work, but training and readiness lag as employers increasingly expect daily AI tool usage.
Q1 2026 cybersecurity funding reached $4.9 billion globally, slightly down sequentially but well above year-ago levels, showing robust investor interest.
This article explains why the EU’s NIS2 directive is a board-level concern for companies, even those not directly affected.
AI’s danger isn’t creating new bugs but amplifying old vulnerabilities.
🎧 Latest Podcasts
🎧 Apr 20, 2026
🎧 Apr 19, 2026
🎧 Apr 20, 2026
