The day's cybersecurity headlines from public sources, in one place. Search them, filter by theme or CVE, or show only vulnerabilities CISA has confirmed are being exploited. Every headline links to the original.
For my own read on what the news means for your business, see Insights.
308 stories match. Clear filters
-
Dolphin X: Windows stealer/RAT with an AI profiler to rank victims
Varonis Threat Labs detailed a new stealer/RAT sold by an actor using the alias "Kontraktnik" that targets 300-plus applications for crypto wallets, .env files, SSH keys, cloud tokens, and DevOps credentials. Its standout feature is an "AI Profiler" that scores infected hosts by app usage…
-
AgentForger: single ChatGPT link could stand up an attacker-controlled AI insider
Zenity Labs found a flaw in OpenAI ChatGPT Workspace Agents where one crafted link could build, authorize, and deploy an autonomous agent inside a victim organization, inheriting the employee's access to Outlook, Gmail, Slack, SharePoint, and Teams. The forged agent polled an attacker inbox every…
-
White House launches "Gold Eagle," an AI-driven clearinghouse for coordinating vulnerability fixes across critical infrastructure
The White House, Treasury, DHS/CISA and the Department of War announced Gold Eagle, a system that consolidates vulnerability findings from government and industry, prioritizes the most consequential flaws, and coordinates remediation across critical infrastructure sectors. It builds on a June 2…
-
Sysdig documents JADEPUFFER, the first ransomware attack run start to finish by an AI agent
Sysdig's threat research team found an LLM-driven agent that broke into an exposed Langflow server via an old, already-patched flaw (CVE-2025-3248), harvested cloud and AI provider credentials, moved laterally into a MySQL/Nacos setup using a default signing key from 2020, encrypted 1,342 config…
-
First documented case of an AI model independently building working browser-native ransomware
Check Point found a Python Flask malware sample, built using DeepSeek, that combines a phishing lure with Chrome's File System Access API to encrypt local files entirely inside the browser, no native payload, browser exploit, or root access required, on both Windows and Android. The sample…
-
Apple patches WebKit bugs found by Anthropic Claude and OpenAI Codex Security, cites AI-driven exploit speed
Apple shipped iOS 26.5.2, iPadOS 26.5.2, macOS Tahoe 26.5.2, and Safari 26.5.2 out of cycle, fixing over three dozen flaws including four WebKit bugs found using AI tooling. Three (CVE-2026-43707, CVE-2026-43716, CVE-2026-43745) are credited to OpenAI's Codex Security; the fourth, CVE-2026-43715…
-
"Gaslight" macOS implant weaponizes prompt injection to derail AI-assisted malware triage; tied to North Korea
Researchers documented Gaslight, a previously undocumented Rust-based macOS backdoor and infostealer assessed with high confidence as the work of North Korea-aligned actors. Its standout feature is an embedded cascade of 38 fabricated "system" failure messages in a Markdown-fenced block, designed…
-
OpenAI expands Daybreak: GPT-5.5-Cyber, Codex Security plugin, "Patch the Planet"
OpenAI expanded its Daybreak cybersecurity initiative on June 23, adding a Codex Security plugin (scans codebases, traces attack paths, builds threat models, generates patches, exports SARIF/CodeQL), broader access to GPT-5.5-Cyber (85.6% CyberGym, 39.5% ExploitGym, 69.8% SEC-bench Pro), a Cyber…