Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

The Safety Problem Nobody Warns You About When You Start Training a Language Model

There's a version of the LLM safety conversation that stays comfortably abstract - AI alignment, existential risk, theoretical failure modes that matter at a scale most organizations will never reach. That conversation is important, but it's not the one most product and technology leaders need to be having right now. The one they need to be having is more immediate and considerably more practical: how the specific decisions made during llm training services directly shape whether the model you deploy is one your organization can actually stand behind.

Latency Lessons From Building a ReAct AI Agent for Agentic Search

Egnyte AI surfaces insights from an organization's documents for regulated industries—life sciences, financial services, architecture, engineering, and construction—and does that within existing permissions and compliance controls. Our AI Assistant is the conversational front door. Ask a question about your documents in plain language, summarise a contract, find the latest version, pull a compliance clause, and get an answer grounded only in the files you're permitted to see.

Best Tools for Securing MCP and LLM Integrations

Shadow IT used to mean employees spinning up unsanctioned software-as-a-service (SaaS) apps that stored company data without approval. Today, shadow MCP and unsanctioned LLM integrations represent the next evolution, and they're more dangerous. Model context protocol (MCP) servers don't merely store data; they act on it, executing code, calling APIs, and accessing internal tools on behalf of AI agents that developers connect with a config file.

Your Firewall Rules Are Drifting Right Now. You Just Can't See It

Firewalls are the single most common source of misconfiguration-related breaches, yet they get changed a hundred times a week and audited once a quarter. This is the network security gap AI attackers exploit first. Endpoint gets the budget. Identity gets the roadmap. The firewall gets changed constantly and reviewed rarely. It is also the control most tied to breaches: 42% of security teams pinned a firewall misconfiguration to a breach or near miss last year, ahead of EDR at 40% and identity at 39%.

How to Secure AI Agents in the Enterprise: A Practical Guide for CISOs

Building guardrails for AI agents sounds like a policy problem but it is actually a data problem. You cannot enforce boundaries on behavior you cannot see. And you cannot govern identity for actors you have not discovered. That dependency chain is what most enterprise security programs miss in 2026, and it is where exposure quietly accumulates. A human employee who mishandles sensitive data creates a containable event. An AI agent with the same permissions creates a different problem.

Why Sensitive Data Detection Is Harder in AI Workflows

Sensitive data used to live in predictable places database columns, known field names, structured rows. That changed when data moved into documents. And it changed again when AI workflows arrived. In this video, we walk through why detecting sensitive data in AI pipelines is fundamentally different from traditional data discovery, and why the old approaches break. We cover the four failure modes that make detection hard in AI workflows.

5 Best Predictive Cyber Intelligence Platforms for Enterprise Security Teams (2026)

Most security tools describe what has already happened. The harder question is what happens next: which exposures an attacker will chain together, and where they will get in. CloudSEK's Global Threat Landscape Report 2025 describes cybercrime as a structured, industrial ecosystem built on stolen credentials, access marketplaces, and coordinated attack chains, and frames the response as a shift from reactive defense toward predictive resilience.

Prompt Injection and the Rise of Agentic Risk

Boxers will often say, the punches that hurt the most aren’t the ones which are thrown with the most force, but the ones they didn’t see coming. I think the same is true in cybersecurity. It’s not the most advanced technically efficient, 0-day utilizing attacks that have the biggest impact, but rather those quiet ones. With no malware or suspicious login at three in the morning from an IP address in a country your company has never done business with. No alert fires.

AI Pentesting Buyer's Guide: How to evaluate AI pentesting vendors

Pentesting made sense when releases happened every few months. A point-in-time assessment could provide an accurate picture of risk for weeks, sometimes months. Today, engineering teams ship continuously. Our State of AI in Pentesting survey of 200 CISOs and 200 engineering leaders, found that 76% deploy significant changes at least weekly, while nearly 40% deploy daily. Yet only 21% validate security on every release. That gap has consequences.

MCP Supply Chain Security: How Malicious MCP Servers Are Infiltrating Enterprise AI Environments

Every enterprise deploying AI agents is building on a foundation of third-party MCP servers they don’t control, can’t verify, and barely track. The security conversation keeps focusing on the model – prompt injection, jailbreaks, hallucinations. That’s the wrong place to look. We’ve covered why that framing falls short elsewhere too – see System Prompts Are Not Security Boundaries. Business Logic Graphs Are.