Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

One of the blockers to moving VMs off vSphere and onto Kubernetes is losing NSX and the protection it provides. Security teams that have spent years building out distributed firewall policy look at Kubernetes and are, quite understandably, alarmed by the flat network and the fact that any workload can reach any other by default. How will they enforce east-west traffic controls? Will they be able to replicate NSX distributed firewall rules with the same granularity?

Last summer we found a critical vulnerability in Teleport, our first in a decade. It was assigned CVE-2025-49825. This vulnerability allowed Teleport SSH certificates, issued to users of a cluster, to sign other SSH certificates which would then be incorrectly accepted by Teleport as valid. Once you can sign your own certificates, you can escalate privileges and bypass authentication controls. This is why we classified it as a critical vulnerability.

Simon Hunt, Chief Product Officer, Securonix For years, security teams have built behavioral models around people. We learned that valid credentials only tell so much.

Two weeks after Hugging Face disclosed an AI-driven intrusion into part of its production infrastructure, Anthropic said it had found three incidents from its own cyber evaluations. Hugging Face traced its compromise through a data-processing pipeline, where a malicious dataset abused a remote-code dataset loader and a template-injection issue in a dataset configuration.

Attackers are getting smarter. AI is making social engineering more convincing, more personalized, and harder to spot than ever before. Training the digital workforce, employees and agents, to recognize threats is necessary, but even the most security-conscious users can still make a mistake at the moment of risk. Workforce risk is a behavior change problem, and effective behavior change requires knowledge, pressure-tested application and real-time reinforcement all working together.

By Shikhar Dalela and Jeewan Singh Jalal The operators named the kit themselves. Buried inside compromised legitimate websites, the hidden staging directory is sometimes literally called “/.bulletproof”, and the PHP session cookie the kit sets on every visitor is named “bp_redir_sess.” The “bp” stands for bulletproof, which is an unusual degree of candor from a threat actor whose entire design philosophy is concealment.

AI governance reduces security risk by enforcing least-privilege access, protecting the data and credentials AI systems handle and making every AI action auditable. This matters because most organizations deploy AI faster than they can govern it. Employees adopt unsanctioned tools, and autonomous AI agents are created under existing user identities. Each one adds unmonitored machine identities that expand your attack surface – the exact gap that governance closes.

Agentic attacks are changing the speed and scale of the defensive challenge. This post explains the challenge, Cato’s new purpose-built capability for addressing it, and what we observed in an end-to-end demonstration.