Microsoft has launched the public preview of Confidential Virtual Machines on Azure Local, enabling organizations to deploy confidential computing workloads on their own infrastructure. This preview allows customers to run sensitive workloads with hardware-based security, bridging the gap for those unable to move critical data to the public cloud. The offering is currently validated on select AMD EPYC-based Lenovo and HPE edge hardware and supports Ubuntu 24.04 guests.

The OpenClaw companion app for Windows, introduced by Microsoft as a step toward agentic computing, provides a graphical interface for managing AI agents. While the app automates much of the initial installation, configuration is still a complex process that includes selecting model providers, integrating chat channels, and managing agent personalities. As well, the app's interface combines a chatbox with various settings panes, making it confusing for newcomers. While it enables connections to messaging platforms like Telegram, the overall experience favors developers and AI enthusiasts rather than mainstream users, with frequent troubleshooting required for even basic agent tasks.

Microsoft plans to release Windows 11 26H2 in the second half of this year as a small enablement package, allowing quick updates for systems on the Germanium platform. Key features include a movable taskbar, Start menu customization, improved search with a local-only option, enhanced Copilot integration, and default settings backup. The update will offer 24 months of support for Home and Pro editions and 36 months for Enterprise and Education.

Cloudflare has open-sourced Cloudflare OS, a platform originally designed for internal use, to enable employees, including nondevelopers, to build applications using AI agents. The platform features a robust security framework that uses fine-grained sandboxing to minimize security risks and data breaches. Cloudflare OS is compatible with various AI models and allows administrators to monitor AI usage.

Cybercriminal gangs are recruiting corporate insiders to facilitate attacks and leak sensitive data, according to a report from TrendAI. The report reveals a structured underground economy on Telegram and hacking forums, where insiders are offered fixed payments or profit-sharing agreements. Industries such as social media, telecommunications, financial services, and logistics are particularly targeted. TrendAI advises companies to monitor workflow exceptions and require dual approval for high-risk transactions to mitigate this growing threat.

LeBron James has joined the Philadelphia 76ers with a two-year, $8 million contract, a 92% pay cut from his previous deal with the Los Angeles Lakers. The move highlights a trend where employees prioritize factors other than salary, such as job satisfaction and well-being. "Everyone who stays great for 20-plus years eventually makes the same shift: They stop optimizing for money and start optimizing for meaning," one LinkedIn commenter says.

US employers announced about 33,500 job cuts in July, the lowest monthly total in two years, according to Challenger, Gray & Christmas data. Layoffs are 41% lower than at this point in 2025, with the technology sector most affected, accounting for more than 30% of job losses.

While AI often grabs the headlines, experts say automation and DevOps are the behind-the-scenes technologies helping businesses keep pace with change. By automating routine IT tasks and creating more reliable technology systems, organizations can improve cybersecurity, speed up software development and build a stronger foundation for AI initiatives. In this guest column for SmartBrief, CompTIA's chief technology evangelist Dr. James Stanger examines DevOps' role in the democratization of IT.

Phishing remains one of the most common and dangerous online scams. Fake emails, text messages and websites trick people into sharing passwords, financial information or other sensitive data. Cybercriminals often create a sense of urgency, such as warning that an account will be locked or a payment is overdue, to pressure victims into clicking malicious links. Security experts recommend using strong passwords, enabling multi-factor authentication and carefully verifying unexpected messages before acting. This guide has more practical steps you can take to protect yourself and your organization.