A critical OpenClaw agent orchestration flaw allowed attackers to hijack autonomous AI agents and manipulate their actions, exposing the security risks in AI middleware.
A GitHub Actions supply-chain attack redirected mutable version tags to attacker-controlled commits, enabling credential theft from CI/CD workflows that relied on unpinned actions.
A step-by-step checklist for verifying whether the compromised Nx Console VS Code extension exposed developer secrets, auditing installed extensions, and rotating any potentially affected cloud credentials.
Akamai’s $205 million LayerX acquisition signals a bigger push into AI-driven enterprise browser security, with implications for extensions, authentication, session handling, and JavaScript-based attack surfaces.
A vulnerability in a popular WooCommerce funnel builder plugin is being actively exploited to inject malicious JavaScript that skims payment details during checkout.
A new benchmark suggests advanced AI models can autonomously craft real browser exploits, compressing exploitation timelines and raising urgent concerns for browser security and offensive AI automation.
Grafana Labs’ stolen GitHub token breach shows how a single compromised CI/CD secret can expose source code, trigger extortion, and put open-source supply chains at risk.
A critical NGINX vulnerability, CVE-2026-42945, is being actively exploited in the wild to crash workers and may enable remote code execution against exposed backends, including Node.js deployments behind NGINX.
A critical Next.js credential disclosure flaw is being chained with an actively exploited WooCommerce funnel builder vulnerability to enable cloud account compromise and checkout skimming attacks.
An actively exploited Microsoft Exchange Server zero-day is being used in global attacks, and administrators should apply Microsoft’s emergency mitigations immediately while awaiting an official patch.
Grafana’s GitHub token breach shows how a CI/CD secret leak can expose source code, enable extortion, and turn pipeline weakness into a supply chain security incident.
A critical Next.js vulnerability can expose server component secrets, cloud credentials, and API keys, making rapid version checks and patching essential for developers.
A technical audit of the Shai-Hulud worm explains how it spreads through npm workflows and exfiltrates developer secrets from GitHub, AWS, and Kubernetes environments.
Google blocked its first AI-assisted zero-day attack while researchers used Claude Mythos to bypass Apple M5 protections and uncover an 18-year-old Nginx bug, underscoring how AI is reshaping both offensive and defensive vulnerability discovery.
Microsoft has confirmed active exploitation of CVE-2026-42897 in on-prem Exchange Server through crafted email attacks, with APT activity targeting exposed unpatched systems including organizations in critical sectors.
CVE-2026-20182 is an actively exploited Cisco SD-WAN controller vulnerability that can grant admin access through a malformed request, requiring urgent patching by network teams.
A technical breakdown of the Shai-Hulud worm and how a compromised npm package can pivot through developer pipelines to steal secrets and reach GitHub, AWS, and Kubernetes environments.
A counterfeit npm package campaign tied to the TanStack incident compromised employee devices at OpenAI and Mistral, exposing internal repositories and highlighting the risks of open-source supply chain attacks.
A large-scale npm supply chain attack weaponized 170 packages to exfiltrate GitHub, AWS, and Kubernetes credentials from developers and organizations.
A technical look at the first known AI-generated zero-day exploit, its detection in a mass cyberattack, and what it means for hardening JavaScript supply chains against automated offensive tooling.
A Composer bug exposed GitHub tokens in CI logs, making immediate patching and token rotation essential for affected pipelines.
A critical NGINX vulnerability dubbed Rift could allow attackers to hijack exposed web servers, making it urgent for DevOps and SRE teams to verify configurations and apply patches.
A supply-chain attack compromised 84 TanStack npm packages, using SLSA-attested malware to target CI credentials and expose weaknesses in build integrity controls.
A practical analysis of how a file upload flaw in Open WebUI can lead to one-click remote code execution and what it means for self-hosted AI interfaces.