Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Design system work follows a well-defined loop: read the ticket, check the Figma spec, find the right component primitives, apply the right tokens, write the Storybook stories, run the tests, open the PR. The steps are consistent enough that when we looked at our design system backlog, we didn't just see a list of tasks; we saw a set of instructions waiting to be executed.

Zero-Shot Learning is a podcast for AI builders, hosted by Nancy Wang, Chief Technology Officer at 1Password, and Dev Tagare, Senior Director and Head of Engineering for Gemini Enterprise & Business at Google. Together, they’ve built and scaled AI systems at the infrastructure and product layers and bring a builder's perspective to every conversation.

In a typical enterprise, non-human identities (NHIs) are thought to outnumber human users by at least 50:1. NHIs are various and include: It is estimated that the NHI: human ratio may have leapt to 144:1 as more AI agents were deployed over the last year. CISOs are already alive to the risks posed by orphaned accounts on their systems. They know that automated rotation is required to revoke privileges as soon as NHIs complete tasks.

Grid is LimaCharlie's agentic AI layer for security teams that want AI operations running across their existing stack right now. Security providers and SOCs need access to AI capabilities without waiting for a migration window, a contract renewal, or a vendor to ship the features they need. Every major security vendor is offering some version of AI. CrowdStrike has Charlotte AI. SentinelOne has Purple AI. Microsoft has Copilot for Security.

For MSPs, patching has evolved beyond a routine maintenance task into a core security service that directly impacts client protection, helping reduce risk, improve operational efficiency, and create scalable recurring revenue opportunities. Unpatched vulnerabilities remain one of the primary exposure vectors for organizations today.

Security teams that deployed legacy DLP years ago built something real. The rules fire. The alerts go out. Compliance boxes get checked. The problem is not that those programs stopped working. It is that the threat moved, and the architecture did not. Agentic AI has introduced a class of data movement that legacy DLP was never designed to govern: autonomous, continuous, multi-step, and operating at machine speed across systems that static rules cannot enumerate in advance.

SophosLabs analysts investigated WantToCry ransomware attacks that involved the threat actors abusing the Server Message Block (SMB) service for initial access and then exfiltrating files to attacker-controlled infrastructure for remote encryption. The detection surface is significantly reduced because WantToCry operates without local malware execution, and there is no post-compromise activity beyond exfiltrating files and rewriting them to disk.

Shai Horovitz is the CEO of JIT, bringing a wealth of experience across business, technology, and security.

This article was originally published in TechRadar Pro. The Iran conflict is serving as an AI testbed for the next era of cyber conflict. Most organizations are watching the tactics and impact unfold with cybersecurity defenses that are simply not prepared for this level of sophistication. Meanwhile, technology leaders are seeing AI as both their biggest opportunity and a major new attack vector.

When the US Center for AI Standards and Innovation (CAISI) asked for public input on securing agentic AI systems, the response was massive: over 500 detailed submissions from Fortune 500 companies, defense contractors, AI startups, and cybersecurity firms. The result is substantial insight into how industry views the regulatory challenges of autonomous AI agents and what they think policymakers should do about it.