Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

AI helped researchers find two critical PHP flaws, and one sat in the code for at least a decade. PHP runs roughly 70% of websites, and attackers were probing on the day the patches shipped, October 21. Dirk Schrader, VP of Security Research at Netwrix, lists four things to do today: update to PHP 8.5.12, 8.4.27, 8.3.36, or 8.2.35 and restart the workers; on PHP 8.1 or older, where there’s no official release, use vendor extended support or take the server offline; check your logs and web roots from before the patch; and limit the account PHP runs under.

Turn on target browsing in Trilio for Kubernetes and look inside a backup target on OpenShift: the backup plans, backups and snapshots it holds. You will: Chapters Check that browsing is on (read-only): oc get targets.triliovault.trilio.io -A -o custom-columns=NAMESPACE:.metadata.namespace,NAME:.metadata.name,BROWSING:.spec.enableBrowsing Recorded on Red Hat OpenShift 4.21 with Trilio for Kubernetes 5.4.1.

As AI agents and non-human identities proliferate, managing secrets and identities is a primary security challenge. Keeper Security CTO and Co-founder Craig Lurey met with Cybercrime Magazine at Black Hat USA 2026 to highlight Keeper’s modern, unified platform, which secures passwords and passkeys, secrets, endpoint privileges and privileged access management.

The shadow AI user is notoriously hard to observe. This elusive creature hunts in incognito tabs and leaves almost no trace on the approved tools list. In this rare footage, a looming deadline draws one employee into the open, and an unapproved AI tool is only a few keystrokes away. Even the stealthiest hunter can be stopped, though. UpGuard User Risk uncovers shadow AI across your workforce and redirects people to approved tools in the moment.

The shadow AI user is notoriously hard to observe. This elusive creature hunts in incognito tabs and leaves almost no trace on the approved tools list. In this episode of Cyber Risk in the Wild, a looming deadline draws one employee into the open, and the pride’s AI policy is put to the test. UpGuard User Risk uncovers the shadow AI and unsanctioned SaaS your teams are using, blocks risky uploads and pastes in the browser, and redirects people to approved AI tools in the moment.