“AI”

ESET finds an increase in malicious AI agents in open-source repos designed to trick developers and steal secrets.

Google's Chrome 150 fixes 27 vulnerabilities, including 13 use-after-free bugs and two critical flaws.

AI Now Institute's 'Friendly Fire' shows autonomous AI coding agents can be tricked into executing attacker code during self-approved scans.

Wiz-discovered GhostApproval symlink flaw lets six AI coding assistants run attacker-controlled code by tricking file approvals.

Cisco Talos finds China-linked UAT-7810 expanding its ORB proxy network using new LONGLEASH malware targeting network devices.

Threat group O UNC 066 uses phone-based phishing to trick employees into registering attacker-controlled Entra passkeys to hijack accounts.

AI gateway compromise led to cryptomining, demonstrating exposure of models, cloud resources, and IAM data via gateway attacks.

Roundcube 1.7 patches zero-click stored XSS via crafted MIME attachments enabling remote compromise without user interaction.

ACSC warns a global campaign is exploiting CMS vulnerabilities to deploy webshells and persist on small to mid-size websites.

State-owned Latvijas Valsts Mezi (LVM) remains recovering weeks after a financially-motivated cyberattack corrupted systems.

Vulnerabilities in cloud-connected baby monitors and cameras exposed nearly one million devices and private video data.

AssuranceAmerica breach exposed nearly 6.9 million drivers' records after attackers accessed internal systems earlier this year.

KDDI data breach exploited a third-party zero-day, affecting roughly 12 million users' email accounts for ISPs.

Sygnia analysis shows agentic AI compressed cloud intrusion timelines to 72 hours, emphasizing speed and orchestration risk.

Researchers detail GodDamn ransomware's use of a signed PoisonX kernel driver to disable security tools and evade detection.