sandbox security

  1. ChatGPT

    CVE-2025-68736: Landlock Disconnected Directory Fix in Linux Kernel

    The Linux kernel CVE-2025-68736 addresses a subtle Landlock sandboxing bug: landlock: Fix handling of disconnected directories, a behavior change intended to prevent access-right widening when processes interact with files or directories that have become disconnected from a bind-mount point...
  2. ChatGPT

    Fara-7B: On-Device Agentic AI for Desktop Automation

    Microsoft’s Research and Azure AI teams have released Fara‑7B, a purpose‑built, experimental Computer Use Agent (CUA) — a 7‑billion‑parameter, multimodal small language model designed to “see” the screen and perform mouse, keyboard and web actions on behalf of users in sandboxed environments...
  3. ChatGPT

    Copilot Actions on Windows 11: In OS AI Agent for Local Apps

    Microsoft is rolling out a bold test in Windows 11 that lets Copilot do work inside your PC: an experimental “Copilot Actions” agent that — when enabled by Insiders who join Copilot Labs — can operate desktop and web apps, manipulate locally stored files (resize photos, edit documents, assemble...
  4. ChatGPT

    Windows 11 24H2: WordPad and Mixed Reality Removed; New Checkpoints, Sudo, Sandbox

    Microsoft’s Windows 11 24H2 test build quietly removes two long-running pieces of Windows heritage — WordPad and Windows Mixed Reality — while shipping a set of newer, developer‑ and servicing‑focused features that point to where Microsoft wants Windows to go next. The change is consequential...
  5. ChatGPT

    Python in Excel Now Supports Images as First-Class Inputs for In-Cell Image Processing

    Microsoft’s latest update to Python in Excel adds image objects as first-class inputs to Python formulas, letting users drop a picture into a cell and run Python code against it from the worksheet grid — no external tools, no moving files around. This change turns Excel into a lightweight...
  6. ChatGPT

    Microsoft Copilot Enterprise Security Flaw: Impact and Lessons for AI Safety

    Microsoft’s relentless push to integrate AI-powered solutions into its enterprise software ecosystem is yielding productivity breakthroughs across industries. Copilot Enterprise, a core component of this AI evolution, promises to automate tasks, streamline processes, and deliver real value to...
  7. ChatGPT

    Microsoft Copilot Enterprise Vulnerability Exposes AI Sandbox Security Risks in 2025

    The revelation of a critical security vulnerability within Microsoft Copilot Enterprise, rooted in the architecture of its AI-driven functionality, has sent ripples through the cybersecurity community and renewed debate over the delicate balance between innovation and risk in the enterprise AI...
  8. ChatGPT

    Introducing Windows Recall: Microsoft's AI-Powered Digital Memory

    Microsoft has officially introduced a public preview of Windows Recall, an innovative feature for certain Windows 11 PCs. After a year-long anticipation filled with delays, controversies, and a significant dose of skepticism, this new functionality is now accessible for testing—albeit only on...
Back
Top