Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

We found more than 16,000 Supabase databases sitting wide open, and over half held personal data like names, phone numbers and passwords. The organizations behind them ranged from a valet service to a government consulate, and in many cases the owners never checked the settings AI wrote. What is the Supabase data exposure? Anyone visiting these sites could read the data in their databases. Many belong to vibe-coded apps, where AI coding tools often handle the database setup.

See CrowdStrike Falcon Data Security for SaaS in action and learn how to discover, classify, and protect sensitive data across Microsoft 365. See how security teams can identify sensitive data in SharePoint and OneDrive, prioritize exposure, automatically apply Microsoft Sensitivity Labels, reduce unintended Microsoft Copilot exposure, and extend protection with just-in-time access.

How do you patch faster - all while honoring your patching policies and *not* breaking things? Attackers are using AI to find and exploit vulnerabilities faster, and vendors are shipping patches faster too. Your patching processes need to keep up! In this episode we show how to automate patching on a monthly cadence, roll it out in rings, and keep control of the process the whole way through. zero-touch, ring-based patch automation with Tanium deployment plans, in Atlas or the classic console.

ShinyHunters breached FBIJobs.gov, leaked personal data on roughly 5,000 Bureau personnel—including 14 staffers tied to China-focused counterintelligence—and triggered arrests in Amsterdam and Jordan. We separate what the FBI confirmed from what the crew claimed: the unverified Oracle PeopleSoft zero-day, the vishing and OAuth token abuse behind 140+ breaches, and why the arrests wound the brand but not the playbook. Plus, the four controls worth validating this week, starting with FIDO2.

Tjeerd Jan van der Molen and Margot Schipper of Cloudflare introduce Cloudflare OS and show how it rewires the way everyone in a company can work with AI.

CrowdStrike's Falcon for XIoT addresses challenges in managing complex industrial and clinical environments by providing visibility into operational assets. The solution helps teams prioritize vulnerabilities based on operational impact rather than severity alone, ensuring safe decisions while minimizing interruptions. Through a detailed investigation process, users can assess asset profiles, validate exposures, and implement controlled responses tailored to specific risks. This approach optimizes asset management and enhances safety in both operational technology (OT) and clinical settings.

Enterprises are deploying AI agents at scale, and those agents are taking real business actions. Through LLMs, MCP servers, and APIs, they move money, access patient records, modify code, and send emails. The attack surface has fundamentally shifted, but most security programs are still focused on what an AI says rather than what it does.

AI is reshaping the IT leadership role faster than any shift in the last decade. As AI makes it dramatically easier to build software (agents, apps, automations, and more), the volume of software entering the enterprise is far beyond what IT teams have had to govern before. That shift demands new ways of thinking about enterprise control: how security, access, and oversight evolve to keep pace with software that's built faster and by far more people than ever before. The leaders who treat this as an operating shift, not just a technical one, are the ones expanding their influence.

Live from Nashville, Joe Smolarski announces big innovations to the WatchGuard Unified Security Platform, including: Agentic AI, Shadow AI, T175, Network management and more.

Attackers don't wait for humans, and neither should your SOC. Tanium Security Operations brings detection, investigation, hunting, and response together on one platform that moves at machine speed. See how Tanium Atlas works as an extension of your team: pulling live telemetry, historical activity, and the full process tree the moment a threat surfaces, hunting a hypothesis across every endpoint in your fleet, and containing threats from one endpoint to hundreds of thousands. When the hunt needs to go deeper, HuntIQ's hands-on experts take it from there, inside your environment.