Author: Carter

Samuel Boivin/NurPhoto via Getty ImagesStay updated with ZDNET: Set us as a preferred source on Google.ZDNET’s main pointsMicrosoft is transforming AI into a tool for prioritizing security threats.Microsoft aims to protect code, AI agents, data, and models.MDASH leverages AI agents to filter out irrelevant security warnings.Last month, Microsoft launched MDASH, known as the Microsoft Security multi-model agentic scanning harness. While the name might not be the catchiest, this initiative represents a significant step forward. Its goal is to streamline the overwhelming flood of security alerts, highlighting only those vulnerabilities that pose a genuine, immediate risk.At Build 2026, the latest update is that…

The foundations of our society are shifting. That was a key insight from Max Buckley’s presentation at AI Engineer Singapore, and it’s been on my mind ever since. For years, software development revolved around limited resources. Writing code was costly, skilled engineers were hard to find, and shipping features required significant time. These constraints dictated how teams operated. We chose priorities carefully because every feature came with substantial trade-offs. Now AI has shattered that old model. With coding agents growing more powerful, the price of building software is plummeting. Tasks that once took weeks can now be mocked up in…

In brief At Build 2026, Microsoft unveiled Scout—its inaugural “Autopilot” agent. Scout is powered by OpenClaw, the open-source agent framework that reached 180,000 GitHub stars within three months of its January 2026 debut. Microsoft is integrating enterprise-grade security into Scout and contributing policy controls to the project. Copilot alone isn’t enough. Microsoft is now building an AI that doesn’t wait to be asked. Announced at Build 2026, Microsoft Scout integrates with your Teams messages, Outlook, OneDrive, and SharePoint, then quietly works behind the scenes to handle the coordination tasks you keep putting off—scheduling meetings across different time zones, highlighting decisions…

The time it takes to exploit security flaws after they’re discovered is getting shorter at breakneck speed—and this trend shows no signs of reversing. Vulnerabilities are being found, recreated, and used as weapons more rapidly than ever before in the history of corporate security. Consequently, the gap between a vulnerability going public and widespread attacks being spotted online is now counted in hours, not days. The standard response from the cybersecurity industry has largely been: apply patches more quickly. Regulatory bodies promote it, company boards anticipate it, and senior executives insist on it. But for the majority of organizations, flipping…

For years, cybersecurity defense worked on a predictable schedule of about two to three weeks. When a security flaw was uncovered, attackers usually needed 15 to 20 days to study the code, build an exploit, and prepare it for use. This delay gave defenders enough time to install patches or reconfigure their network protections. That breathing room has vanished. With nation-states now embedding artificial intelligence and generative AI into their attack methods, the gap between discovering a weakness and exploiting it has shrunk from weeks to mere hours, often happening before the vulnerability is even made public. In this transformed…

NVIDIA has unveiled its Factory Operations Blueprint, a reference architecture for creating autonomous factory systems.Factories typically rely on separate, siloed systems. Programmable Logic Controllers (PLCs) manage machine automation at a low level. Supervisory Control and Data Acquisition (SCADA) systems monitor processes. Manufacturing Execution Systems (MES) track production workflows. Enterprise Resource Planning (ERP) software handles business operations. These systems often struggle to work together, limiting overall plant intelligence and blocking the use of advanced AI for predictive or proactive maintenance.Without a complete view of operations, finding the cause of production delays is a slow, manual task. Quality checks often depend on…

Adhesive bonding is an assembly method favored by engineers for creating robust joints. However, industrial adhesives come with certain weaknesses—notably outgassing—which can create complications in some applications. When materials are placed in a vacuum or low-pressure environment, adhesives may emit gases that can taint nearby surroundings or disrupt sensitive machinery. This issue is especially problematic in aerospace applications, along with products that depend on optical components, optoelectronics, or photonics. Semiconductors and medical imaging tools—endoscopes, for example—are also vulnerable to these difficulties. Substances released through outgassing can corrode and damage fragile electronic circuits. They can also haze over lenses and lead…

Anthropic’s decision to file for an IPO signals a major shift for generative AI: it is evolving from an experimental, research-driven phase into a reliable, everyday tool for businesses.While operating privately, AI developers have focused mainly on building faster and maximizing computing power, often ignoring predictable billing. Taking a leading AI company public forces its engineering priorities to align with standard business purchasing needs, bringing in regular release dates and clear pricing structures that corporate leaders need for long-term planning.William Samengo-Turner, Technology Sector Lead at A&O Shearman, commented: “If Anthropic goes through with an IPO, the biggest question isn’t whether…

  # Introduction  Over recent years, AI Explainability (XAI) has become a major focus in the world of real-world AI systems, and Large Language Models (LLMs) are no exception. Because these models are so complex and powerful, it is crucial to shift from static testing to dynamic evaluation. This helps us better grasp how these mysterious “black-box” systems produce their text outputs. Additionally, blending dynamic evaluation with solid statistical methods and cost-effective, production-ready monitoring tools is becoming a quietly significant trend in the industry. This piece explores the concept of LLM explainability. It highlights the latest advancements, trends, and ongoing work…

Bitcoin sits in an interesting gray area — part commodity, part currency, part tech asset, part macro hedge. That’s not just some abstract thought experiment; it’s the single biggest factor that determines how bitcoin behaves in the market.Since investors haven’t yet agreed on what bitcoin fundamentally represents, there’s no unified rulebook for how it should trade. Every group of buyers brings its own lens and story about what bitcoin means, turning the market into a lively tug-of-war between competing ideas. That push-and-pull, more than anything else, drives bitcoin’s price.In reality, the most powerful group — macro and institutional money —…