Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Critical vLLM Flaw Exposes the Soft Underbelly of AI Infrastructure

While the world worries about "jailbreaking" LLMs or preventing them from hallucinating, a critical new vulnerability has just reminded us of a fundamental truth: AI is just software, and software has bugs. A newly discovered critical flaw (CVE-2025-62164) in vLLM, one of the most popular libraries for serving large language models, allows attackers to achieve Remote Code Execution (RCE) or crash servers simply by sending a malicious API request. This isn't a failure of the AI model.

Beyond security theater: How automated trust closes the AI readiness gap

‍ AI is transforming businesses at breakneck speed—but security isn’t keeping up. ‍ According to Vanta’s State of Trust Report 2025, which surveyed over 2,500 business and IT leaders around the world, 3 in 5 say AI-related security threats are outpacing their expertise. With a majority of organizations experiencing threats weekly, AI is not just driving the volume, but the precision of these attacks.

Why Granular Backup And Recovery Are Essential for your DevOps backup strategy

Every IT stack may look tidy on a diagram. If so, then it’s tempting to assume everything works fine. And yet, systems rarely fail as a whole. Usually, it’s a part or functionality. For instance, anyone who ever untangled a broken workflow in GitHub, GitLab, Bitbucket or Azure DevOps, or a corrupted field in Jira, knows it too well. And that’s the quiet tension (“to fix one little thing”) inside every modern backup strategy.

A Deep Dive Into ggshield, The GitGuardian CLI

In this in-depth walkthrough, we will show you how to turn ggshield, the GitGuardian CLI, into a practical guardrail for keeping secrets out of your code and CI pipelines. You’ll see exactly how to install and authenticate ggshield, then use it to scan repositories, local paths, archives, Docker images, PyPI packages, and CI environments for hardcoded credentials. We’ll also walk through configuring Git hooks with ggshield install.

Ep 3. Palo Alto Networks Research and AI Generated Attacks

AI isn’t just changing cyber defense—it’s transforming how attacks happen. In this episode, Tova Dvorin sits down with Tomer Bar and Shelly Zucker from SafeBreach to explore how AI-powered threats are reshaping the battlefield. Hear about: If you’re in cybersecurity, risk, or IT, this is your must-listen guide to preparing for the AI-enabled threat era.

Ep 4. ToolShell in the Wild: SharePoint Zero-Day CVE-2025-53770 Explained

In this special episode, host Tova Dvorin sits down with SafeBreach experts Adrian Culley and Tomer Bar to unpack CVE-2025-53770 — a zero-day deserialization flaw in Microsoft SharePoint Server that enables unauthenticated remote code execution and long-term persistence. This isn’t theoretical. It’s actively exploited and tied to the evolving ToolShell attack chain. Here’s what you’ll hear in this episode.

Ep 2. FBI Advisory, Iranian Threats & Resilience

The FBI, NSA, and CISA just issued a warning about Iranian state-backed actors, including the notorious Cyber Avengers, targeting US networks—especially OT, IoT, water, and aviation systems. These groups aren’t hacktivists—they’re highly skilled, sanctioned members of the IRGC. Key takeaways: Stay proactive: run simulations, remediate vulnerabilities, and lock the stable door before the horse bolts.

Ep 5. Interlock Ransomware: Don't Accept Code from Strangers

In this episode of the SafeBreach Cyber Resilience Podcast, host Tova Dvorin and Adrian Culley dive deep into Interlock—one of today’s most aggressive ransomware operations. What you’ll learn: From hospitals to schools, no one’s immune—hear how Interlock is rewriting the ransomware playbook and what your team can do to stay resilient.

Ep 6. Storm-2603 & Warlock: Where Ransomware-as-a-Service Gets Real

A new breed of ransomware is here — and it’s more dangerous than ever. In this episode of the Cyber Resilience Podcast, we unpack the chilling rise of Warlock ransomware, a campaign tied to Chinese threat actor Storm-2603. Discover how this group is combining nation-state tactics with ransomware-as-a-service operations, blurring the line between espionage and profit—and what it means for critical infrastructure defense.