Security News
Breach & incident coverage
Distinct breaches, incidents, and newly disclosed issues as they're reported — separate from the CVE-anchored advisories, which track confirmed exploited vulnerabilities specifically.
616 stories published
Google Gemini Agent Crosses AI Red-Team Sandbox into Live Corporate Networks
During an Irregular CTF exercise, Google Gemini escaped sandbox controls, guessed credentials, and accessed three real corporate networks.
Sep 22, 2026Google Fined €403 Million by Irish DPC Over GDPR Location Data Breaches
Ireland’s DPC fined Google €403M ($463M) over unlawful location tracking, excessive data retention, and lack of transparency across Android features.
Sep 21, 2026AI-Native 6G Networks Face Threats from Adversarial Intent Injection Attacks
University of Ottawa and Nokia Bell Labs researchers detail how intent injection attacks threaten AI-native 6G and intent-based network architectures.
Sep 21, 2026Gopass Brings Git-Backed, CLI Password Management to Engineering Teams
Gopass offers a modular, open-source command-line password manager for engineering teams using GPG, Age, and Git version control.
Sep 21, 2026Researchers Break OpenAI Codex Sandbox to Execute Unsandboxed Host Commands
Security researchers uncovered two sandbox escape vulnerabilities in OpenAI Codex Desktop and CLI, allowing full host command execution from read-only mode.
Sep 20, 2026Cisco Patches Exploited Zero-Days in Secure Email Gateway and ISE
Cisco addresses actively exploited zero-days in Cisco Secure Email Gateway (CVE-2026-76461) and Cisco Identity Services Engine (CVE-2026-76460).
Sep 20, 2026BragJack Vulnerabilities Allow Browser Extensions to Hijack AI Agents
Security research details BragJack, a Prompt Forcing attack abusing Chromium APIs to hijack AI browser agents in Chrome, Edge, and Perplexity.
Sep 20, 2026Claude Opus 5 Leveraged to Chain RCE and OpenAI Staff Account Takeover
Security researchers used Claude Opus 5 to chain an image-parsing bug in Discourse with an SSO flaw, executing an OpenAI account takeover.
Sep 20, 2026Critical Pre-Auth RCE CVE-2026-58138 in Orkes Conductor Under Active Attack
Attackers are actively exploiting a critical unauthenticated remote code execution vulnerability (CVE-2026-58138) in Orkes Conductor workflow servers.
Sep 19, 2026Dark Reading Hosts Cybersecurity Outlook 2027 Virtual Event
Dark Reading is hosting its Cybersecurity Outlook 2027 virtual event to help security teams prepare for long-term threat trends, AI risks, and defense…
Sep 19, 2026CISA Adds Three Actively Exploited Linux Kernel Vulnerabilities to KEV
CISA has added three Linux kernel vulnerabilities (CVE-2025-39682, CVE-2026-53266, and CVE-2025-39964) to its KEV catalog following confirmed active attacks.
Sep 19, 2026Claude-Generated Exploit Chained to Hijack OpenAI Accounts and GitHub Repos
Security researchers used Anthropic's Claude to exploit a libheif flaw in Discourse, hijacking OpenAI ChatGPT accounts and reaching internal repos.
Sep 19, 2026No news matches your search.