ITOps Times

SAN FRANCISCO, CA — Cribl, the AI Platform for Telemetry, today announced it is acquiring CardinalOps, an Agentic Detection Engineering solution. The acquisition adds detection engineering capabilities that help customers improve threat coverage, lower data costs, and strengthen their SOCs. This creates a flexible path toward replacing legacy SIEM architectures.

Most production AI agents make authority decisions the same way a firewall does. A request arrives, a policy is checked, and the action is either allowed or denied. That binary worked for decades because the underlying assumption held: if the system made a wrong call, you could undo it. A bad database write gets rolled back. An erroneous permission grant gets revoked. The cost of a mistake was recoverable, so a yes-or-no gate was enough.

Communication failures can create uniquely complicated logistical problems for businesses. Unlike a downed payment gateway or a front-end glitch, which causes measurable revenue loss and can easily be traced when service resumes, a delayed authentication code or a notification that never lands doesn’t register as a single outage on a dashboard. Technical issues often surface later as support tickets that are hard to trace back to their actual cause.

As organizations face growing complexity across networks, cloud environments, applications, AI workloads and digital services, they require deeper visibility, actionable intelligence and greater operational control. Further, with AI taking on greater responsibility, the quality and completeness of the data it acts on has become mission critical.

Despite the shift toward SaaS and managed platforms, many modern applications still rely on components that run inside customer-controlled environments. These self-hosted agents extend functionality and enable powerful integrations, but site reliability engineers (SREs) must still maintain the operational reliability of applications deployed in infrastructure they do not fully control. The model has clear advantages.

An AI agent logged into a production database at 2:47am. It was supposed to clean up stale records. Instead, it deleted 340,000 rows from a table it wasn’t supposed to touch. The incident took six engineers 11 hours to diagnose and reverse. Not because recovery was technically hard. Because nobody could tell them, definitively, what the agent had actually decided and why. That’s agentic AI’s real production problem. Not capability. Visibility.

The utility sector is currently undergoing a massive capital transformation to support grid modernization and distributed energy resources. Duke Energy is executing a $145 billion capital project plan, and Ameren is driving a $1.6 billion transmission expansion program. Operating a modernized grid of this scale requires running hundreds of millions of grid-planning simulations across hybrid environments to forecast demand and manage energy loads.

Recent large-scale disruptions, including regional cloud service degradation in the Middle East and widespread internet routing failures observed in late 2025, have reinforced a critical reality: failures are no longer confined to individual services or availability zones. Entire regions can become partially or fully unreachable with little warning.

Despite massive investments in cybersecurity tools, most organizations are struggling to defend themselves effectively, according to a new global study recently released by threat management company Filigran. The State of Threat Management Report, based on a survey of 550 security leaders, reveals that while companies have plenty of data, they are failing to translate that information into fast, effective action.

The integration of artificial intelligence into software development is bringing a seismic shift to the tech industry, especially when it comes to governance and security in the supply chain. Industry experts agreee that the industry has moved from a “finding the issue” problem to a risk management problem—and the landscape has changed permanently.