Fastly launches AI Runtime Control and AI Firewall for enterprise AI security
New Fastly capabilities aim to give organizations real-time visibility and control over AI models and agents.
Breaches, offensive and defensive AI, and surveillance.
New Fastly capabilities aim to give organizations real-time visibility and control over AI models and agents.
Attacks on Colorado water utilities disrupted operations by altering equipment settings and disabling alarms, highlighting risks to critical OT infrastructure.
A new incident report reveals that even advanced models like Claude struggle with basic human verification puzzles.
As some large firms plan to run SAP ECC beyond 2027, the decision to pay for extended support until 2030 carries costs and security implications.
A weekly roundup of security stories that may have been overlooked, including a ransomware developer's sentence, an AI-powered attack on a WordPress plugin, and a critical SAP vulnerability.
A weekly security roundup highlights an urgent Cisco fix for an actively exploited zero-day and new details on a Revolut breach that used a spoofed government email.
A campaign uses SEO-optimized GitHub repositories posing as trusted software to deliver a previously unknown credential-stealing malware.
A BleepingComputer explainer highlights how Microsoft 365's easy sharing can leave sensitive data accessible long after it should be, and argues owner-driven access reviews restore visibility.
A new UN report says Vietnam, Laos, Pakistan and Argentina have taken meaningful steps against North Korean IT worker schemes.
An open-source password manager positions itself as a drop-in replacement for pass, targeting teams that work in the terminal.
A new survey shows that legacy workflows are driving most AI governance problems at large companies.
Multifactor authentication remains vital, but it cannot substitute for proper OAuth governance and consent oversight.
A Pakistan-aligned group is targeting government and defense entities with a new Rust-based backdoor that uses private GitHub repositories for command-and-control.
A newly documented Windows privilege escalation flaw, CVE-2026-66804, shows how an incomplete patch for the 'Dark Elevator' bug left a dangling COM object registration attack surface.
Researchers warn that the shift to intent-based networking in 6G could open a new attack surface for malicious intent injection.
Admins will soon be able to customize which file extensions Teams treats as security threats.
The update strengthens the Java gateway's role as a bridge for users who want to drop Outlook without losing access to Exchange mailboxes.
A new CSC report identifies impersonation, phishing, and domain abuse as the most pressing online intellectual property risks.