🤖

AI Security

26 articles in this category

AI Security18 min readJun 25, 2026

Microsoft Copilot for Security: Six Months In, What Actually Works

Your SOC team activated Copilot for Security six months ago expecting AI-driven incident response. Some capabilities delivered real analyst time savings. Others produced confident-sounding summaries that were factually wrong. This review covers what Copilot actually accelerates in production SOC workflows, what still requires heavy prompt engineering, and where the token economics make it hard to justify at scale.

Microsoft Copilot for SecuritySOCSentinel

🤖

AI Security18 min readJun 23, 2026

Secure AI Supply Chain: Verifying Models Before Deploying to Azure AI Foundry

A data scientist pulled a community model from the Foundry catalog and deployed it to a production hub. The model contained a pickle deserialization payload that executed under the hub managed identity, giving the attacker access to Key Vault and connected storage. This guide covers model provenance verification, automated scanning pipelines, registry hardening, and the Azure Policy controls that prevent untrusted models from reaching production compute.

AI Supply ChainAzure AI FoundryModel Security

🤖

AI Security19 min readJun 23, 2026

OWASP LLM Top 10 2025: What Changed and What It Means for Azure AI Deployments

The OWASP LLM Top 10 2025 revision reshuffled the risk landscape: prompt injection dropped to second place, unbounded consumption is new, and system prompt leakage got its own category. If you run Azure OpenAI or AI Foundry workloads, every change maps to specific controls you either have or are missing. This guide breaks down each updated risk with Azure-native mitigations, detection queries, and the controls that actually close the gaps.

OWASP LLM Top 10Azure AI SecurityPrompt Injection

🤖

AI SecurityJun 21, 2026

What Is an AI Firewall? Runtime Protection for Enterprise AI in 2026

AI firewalls inspect prompts and outputs in real time to block injection attacks, PII leakage, and jailbreaks that WAFs and NGFWs are blind to. Here is what they do, how leading vendors approach the problem, and how to evaluate one for your enterprise stack.

AI SecurityEnterprise SecurityRuntime Protection

🤖

AI Security17 min readJun 18, 2026

Prompt Injection in Enterprise AI Copilots: Detection and Prevention

An attacker embedded a hidden instruction in a SharePoint document, and M365 Copilot followed it during a summarization request, exfiltrating internal project names to an external URL. Prompt injection is the top attack vector against enterprise copilots, and most organizations have no detection in place. This guide covers the attack taxonomy, Azure AI Content Safety prompt shields, Defender for Cloud Apps policies, and the KQL queries that catch injection attempts in audit logs.

Prompt InjectionAI SecurityM365 Copilot

🤖

AI Security8 min readJun 7, 2026

The Hidden Risk of AI Skills and MCP Servers: What to Check Before You Install

Installing a Claude Code skill or MCP server takes 30 seconds. Auditing one properly takes longer. With 36% of published skills containing security flaws and documented supply chain attacks already in the wild, here is what to inspect before you run anything.

AI SecurityClaude CodeMCP

🤖

AI Security9 min readJun 6, 2026

AI Security Mistakes You Are Probably Making Right Now

From sharing API keys in chat to installing unvetted browser extensions, the most dangerous AI security mistakes are the ones that feel routine. This guide covers the six most common missteps, with real incident data and practical fixes for each.

AI SecurityChatGPTAPI Keys

🤖

AI Security14 min readMay 28, 2026

Google Agent Garden Cyber Guardian: Build an AI SOC Agent with ADK

Google Agent Garden includes cybersecurity-oriented ADK samples and reference patterns for multi-agent security operations. This article uses the Cyber Guardian pattern to explain alert triage, log investigation, threat intel correlation, and playbook-driven response recommendations without treating the sample as a production SOC replacement.

Google ADKAgent GardenCyber Guardian

🤖

AI Security18 min readMay 21, 2026

Azure AI Foundry Evaluation Security: Adversarial Testing and Red Team Workflows

Content filters and manual review will not catch indirect prompt injection via poisoned RAG documents or multi-turn jailbreak escalation. This guide covers the full operational red team workflow for Azure AI Foundry: PyRIT setup, orchestrator-driven attack campaigns, Azure AI Evaluation SDK safety gates, CI/CD integration, and KQL detection for production probing.

Azure AI FoundryRed TeamPyRIT

🤖

AI Security17 min readMay 19, 2026

Microsoft Purview for AI Governance: Classifying and Protecting AI Training Data

AI training pipelines bypass traditional DLP controls because they access data as bulk blob reads, not document downloads. This guide shows how to configure Microsoft Purview specifically for AI data scenarios: scanning training datasets, designing a label taxonomy for AI use cases, enforcing DLP policies against AI pipelines, and integrating with Azure AI Foundry.

Microsoft PurviewAI GovernanceData Classification

🤖

AI Security22 min readMay 17, 2026

MCP Server Hardening Case Study: Locking Down a Corporate Dev Environment

Most teams treat MCP servers as developer tooling. They are infrastructure, and the incident logs prove it. This guide walks through network isolation, authenticated gateways, Azure Policy governance, and KQL detection for enterprise MCP deployments, drawn from a real post-incident remediation.

MCPClaude CodeAI Security

🤖

AI Security20 min readMay 7, 2026

Azure AI Foundry Security: Threat Model, RBAC, and Data Governance Controls (2026)

Azure AI Foundry introduces hubs, projects, and layered managed identities that fundamentally shift your Azure security model. This guide covers six critical threat scenarios: from cross-team data exfiltration to MI lateral movement, with correct RBAC design, data governance controls, and KQL queries for detection.

Azure AI FoundryAzure SecurityRBAC

🤖

AI Security18 min readApr 26, 2026

Azure AI Foundry Private Link Setup: Secure Azure OpenAI, AI Search, and Storage End-to-End

Securing Azure OpenAI alone is not enough if Azure AI Search, Storage, or Key Vault still expose data over public paths. This guide shows how to build an end-to-end private Azure AI Foundry architecture using Private Link, Private DNS, and segmented subnets.

Azure AI FoundryAzure OpenAIAzure AI Search

🤖

AI Security18 min readApr 18, 2026

MCP Server Security: How to Protect AI Agents from Prompt Injection and Tool Abuse (2026)

Model Context Protocol (MCP) servers are RSAC 2026's hottest security topic. As 40% of enterprise apps embed AI agents by year-end, MCP is the attack surface no one is talking about. This guide covers prompt injection via tools, server impersonation, privilege escalation, and the controls that actually stop these attacks.

MCPModel Context ProtocolAI Security

🤖

AI Security16 min readMar 9, 2026

Non-Human Identities (NHI): The Hidden Security Crisis Powering AI Agent Attacks in 2026

Machine identities now outnumber humans 40–100:1 in enterprise environments. With AI agents minting thousands of new credentials daily, NHIs have become the fastest-growing and least-governed attack surface in cybersecurity. Here is what every security team needs to know.

Non-Human IdentitiesNHIMachine Identities

🤖

AI Security15 min readMar 3, 2026

AI Red Teaming: How to Test Your AI Systems for Security Vulnerabilities

AI red teaming is the practice of proactively testing AI systems for security vulnerabilities and unsafe behaviors. Learn the methodology, tools like PyRIT and Garak, and how to integrate AI red teaming into your secure SDLC.

AI Red TeamingLLM SecurityAI Testing

🤖

AI Security14 min readFeb 15, 2026

How to Secure Your OpenAI and Claude API Integration

Most AI applications ship with exposed API keys, no rate limiting, and zero input validation. Here is the practical checklist for locking down your LLM API integration before something goes wrong.

AI SecurityOpenAIClaude

🤖

AI Security16 min readFeb 12, 2026

The Four Attack Surfaces of AI Systems: Network, Prompt, Data, and Model

AI introduces attack surfaces that traditional security tools were not built to handle. Understanding these four layers - and their distinct threats - is the foundation of any serious AI security strategy.

AI SecurityPrompt InjectionLLM Security

🤖

AI Security14 min readFeb 10, 2026

Microsoft Security Copilot: Complete Guide for Security Teams in 2026

Microsoft Security Copilot integrates AI into every layer of your security operations. Learn deployment, top use cases, and how it changes day-to-day work for security analysts and architects.

Microsoft Security CopilotAI SecuritySOC

🤖

AI Security17 min readFeb 8, 2026

On-Premises AI Security: Protecting Self-Hosted LLMs and GPU Infrastructure

Running AI on your own infrastructure gives you control over your data. It also means you own the security. Here is how to secure Ollama, vLLM, and other self-hosted AI deployments properly.

On-Premises AIOllamavLLM

🤖

AI Security15 min readFeb 5, 2026

Public Cloud AI Security: Azure OpenAI, AWS Bedrock, and Google Vertex AI

Cloud AI services come with strong security capabilities built in. Most breaches happen because those capabilities are never configured. Here is what to configure on each major platform.

Azure OpenAIAWS BedrockGoogle Vertex AI

🤖

AI Security13 min readFeb 1, 2026

AI Security in 2026: What Every Professional Needs to Know

AI security is becoming its own discipline. Whether you are a security professional, a developer deploying AI, or a leader making decisions about AI adoption, here are the fundamentals that matter.

AI SecurityLLM SecuritySecurity Fundamentals

🤖

AI Security28 min readJan 30, 2026

OWASP Top 10 for Agentic AI Security 2026: Complete Enterprise Implementation Guide

The OWASP Top 10 for Agentic Applications 2026 defines critical security risks for autonomous AI agents. Learn how to protect your enterprise from prompt injection, rogue agents, and tool misuse with practical implementation strategies.

Agentic AIOWASPAI Security

🤖

AI Security15 min readJan 26, 2026

How to Secure Azure OpenAI Network Traffic: A Private Endpoint & Terraform Guide

Exposing Azure OpenAI via public networks is a security risk for enterprise data. Learn how to build a fully private architecture using Azure Private Link, disable public access, and deploy it all via Terraform.

Azure OpenAIPrivate EndpointTerraform

🤖

AI Security13 min readJan 5, 2025

AI Security: Risks You Need to Know and How to Mitigate Them

As AI tools become common in enterprises, so do the security risks. Learn about prompt injection, data leakage, and how to use AI safely in your organization.

AI SecurityLLMPrompt Injection

🤖

AI Security18 min readJan 2, 2025

Automating Incident Response: How AI Can Help Your SOC

Security teams are overwhelmed with alerts. Learn how AI and automation can help triage incidents, reduce response times, and let analysts focus on real threats.

SOCIncident ResponseAI

Explore Other Categories

⚙️Cloud Engineering(1)☁️Cloud Security(40)🔐Cybersecurity(13)📝Identity Security(4)🌐Networking(2)📝Security Tools(1)🛡️Zero Trust(10)