Is Microsoft Defender crashing on your PC? A bad update is breaking scans

windowscentral.com Aug 20, 2026

Is Microsoft Defender crashing on your PC? A bad update is breaking scans

Curated by Shubham · TechCapsules · Aug 20, 2026

AI-summarised brief · reviewed before publication

Microsoft Defender, the built-in security application for Windows PCs, is experiencing widespread crashes following a recent software update. Users report that the malfunction specifically disrupts scanning processes, rendering the antivirus protection ineffective during critical operations. The issue stems from a defective update that introduced instability into the core scanning engine. Affected systems are unable to complete routine or on-demand scans, leaving devices potentially vulnerable to threats. Microsoft has acknowledged the problem and is working on a resolution. The bug impacts a significant portion of the Windows user base, causing frustration and security concerns. Reports indicate that the crashes occur immediately when scans are initiated. The company is expected to release a patch to restore functionality. Until the fix is deployed, users may need to disable automatic scanning or rely on alternative security measures. The incident underscores the risks associated with automated update deployments. Technical support channels are currently flooded with reports of the same error. Microsoft aims to minimize downtime and restore full security capabilities as quickly as possible. The situation highlights the importance of rigorous testing before releasing security updates.

💡 Why It Matters

· A broken native antivirus leaves millions of Windows users exposed to malware during the outage.
· This failure erodes trust in Microsoft's ability to deliver reliable security patches without introducing critical vulnerabilities.

Read full article on windowscentral.com

Rust Supply Chain Attack Puts Build-Time Malware in Crates with 245 Million…

The Rust Project removed malicious releases of three popular crates—arrayref 0.3.10, internment 0.8.7, and append‑only‑vec 0.1.9—after a compromised maintainer account published versions that added a typosquatted dependency, proc‑macro1, whose build script fetched and executed remote payloads during compilation. The attacks were detected and the crates deleted within 86 to 107 minutes. Developers are advised to purge cached files, pin arrayref to 0.3.9 or earlier, and monitor for similar build‑time threats.

💡 The incident exposes how build scripts can silently deliver malware, undermining trust in open‑source ecosystems and highlighting the need for stricter credential security and supply‑chain verification.

microsoft.com · Aug 20

Frost & Sullivan names Microsoft a visionary leader in cloud workload protection…

Frost & Sullivan has designated Microsoft as a visionary leader in its 2026 Cloud Workload Protection Platforms (CWPP) report. The analyst firm benchmarked 20 vendors from over 45 qualified candidates, highlighting a market shift toward unified runtime security models. Microsoft earned this recognition for balancing innovation with growth, leveraging its Defender for Cloud framework. The company holds the largest CWPP market share, estimated at over 22% globally. Frost & Sullivan noted Microsoft’s breadth across infrastructure, identities, and applications, which reduces operational complexity for modern and AI-native workloads. The report emphasizes that leadership is now defined by runtime telemetry depth and SOC integration rather than simple scanning. With 82% of container users running Kubernetes in production, runtime visibility is critical. Frost & Sullivan projects CWPP spending to rise from $6.43 billion in 2025 to $7.95 billion in 2026, growing at 19.1% annually through 2030. This growth reflects organizational priorities to secure cloud-native architectures, meet regulations, and protect complex environments involving containers, serverless functions, and AI services through consolidated, context-rich security platforms.

💡 The designation validates the industry pivot from static vulnerability scanning to dynamic runtime protection as the primary defense mechanism. Microsoft’s dominance confirms that unified platforms addressing identity and container sprawl are becoming essential infrastructure for enterprises managing complex, AI-driven cloud environments.

analyticsinsight.net · Aug 20

ChatGPT Could Soon See Everything You Do: Productivity Boost or Privacy Risk?

OpenAI has launched Computer History, a new Mac feature enabling ChatGPT to record user activity across selected applications and websites. Arriving in August 2026, the system logs clicks, typing, keyboard shortcuts, and app switches to create a detailed timeline. This data allows the AI assistant to recall past work, locate files, and resume unfinished tasks with enhanced context. Unlike previous models, Computer History does not capture screenshots, video, or audio, focusing instead on event-based records. While designed to boost productivity through personalized automation and improved task management, the feature raises significant privacy concerns. Critics warn that aggregated activity data can reveal detailed patterns regarding work habits and personal interests, creating a comprehensive digital profile. Security experts also highlight potential risks, noting that increased AI access to computer history could expose users to prompt injection attacks and other vulnerabilities. The feature represents a major shift in AI capabilities, moving beyond simple chat interactions toward deeper system integration and memory retention.

💡 The shift from conversational AI to persistent behavioral tracking fundamentally alters the user-assistant dynamic, transforming software from a passive tool into an active observer of daily routines. This evolution forces a critical reevaluation of digital consent, as the granular data collected for convenience inherently constructs a detailed psychological and professional profile of the user.

forbes.com · Aug 20

How Shadow IT Threatens Your Cybersecurity And Digital Sovereignty

Unauthorized software usage, known as shadow IT, has evolved from a minor hygiene issue into a critical enterprise risk threatening cybersecurity and digital sovereignty in 2026. Employees increasingly adopt unsanctioned applications, artificial intelligence tools, and cloud services, often transferring sensitive data and operational control outside established corporate boundaries. This trend undermines fundamental corporate IT security tenets by creating unmonitored vulnerabilities. To mitigate these escalating threats, IT and human resources departments must collaborate effectively. Without such unified action, the rapid adoption of unauthorized AI tools could transform manageable annoyances into significant security breaches. The resulting sovereignty gap places control over data and infrastructure at risk, directly impacting organizational resilience. Addressing this requires immediate strategic intervention to prevent unauthorized systems from compromising the integrity of digital operations and exposing companies to severe cyber threats that bypass traditional security protocols.

💡 The convergence of AI adoption and shadow IT creates an immediate vulnerability where data sovereignty is lost before security teams detect the breach. Organizations that fail to align IT and HR strategies risk ceding control of their most sensitive assets to unregulated external platforms.

helpnetsecurity.com · Aug 20

8,539 reasons to rethink how vulnerabilities get patched

Rapid7’s Q2 2026 Threat Landscape Report documented 8,539 high‑ and critical‑severity vulnerability disclosures, double the count from a year earlier, underscoring a tightening window for remediation. Exploit code appears faster, with a 76 % rise in publicly available proofs of concept compared to Q2 2025, and 62 % of newly exploited flaws are network‑exploitable, requiring no authentication or user interaction. The report warns that reliance on periodic patch cycles and CVSS scores alone leaves organizations chasing “ghosts” while attackers automate kill‑chains and weaponize patches almost instantly. It recommends shifting focus from sheer CVE volume to exposure‑based risk, maintaining accurate inventories of internet‑facing assets such as VPNs and routers, and enforcing authentication on exposed endpoints to curb the expanding attack surface.

💡 Organizations that continue treating patching as a ticket‑closing exercise are effectively funding attackers’ research, while a exposure‑centric approach can cut the path to compromise.

digit.in · Aug 20

OpenAI tests new AI safety system to spot cyber threats while keeping…

OpenAI has begun testing a new safety feature called Private Safety Processing, designed to identify cyber‑threat indicators across multiple AI interactions while preserving customer data confidentiality. The system, currently being piloted with early partners such as Microsoft and Databricks, analyzes patterns that emerge only when conversations are examined collectively, rather than in isolation. OpenAI says this approach can spot coordinated attempts to extract vulnerabilities, like sequential queries about software weaknesses and remote‑access tools, that would be missed by single‑prompt checks. The company plans to roll the feature out broadly and release a technical paper in September, aiming to bolster security for enterprises that rely on advanced language models for sensitive tasks.

💡 By detecting coordinated malicious queries without exposing proprietary data, OpenAI gives businesses a practical safeguard against AI‑enabled espionage and sabotage.