<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>CertiSafe AI — AI Security Articles</title>
  <link>https://certisafeai.com/articles/</link>
  <atom:link href="https://certisafeai.com/feed.xml" rel="self" type="application/rss+xml"/>
  <description>Practical tips for securing chatbots, AI assistants and LLM features.</description>
  <language>en-us</language>
  <item>
    <title>Overwhelmed and Don't Know Where to Start? 5 Simple Steps to Secure AI in Your Small Business</title>
    <link>https://certisafeai.com/articles/ai-security-small-business-where-to-start/</link>
    <guid>https://certisafeai.com/articles/ai-security-small-business-where-to-start/</guid>
    <pubDate>Thu, 19 Nov 2026 12:00:00 GMT</pubDate>
    <category>Getting Started</category>
    <description>A plain-English starting point for small businesses using AI. Five simple steps to protect your data and customers, with no security team or big budget needed.</description>
  </item>
  <item>
    <title>What Is Shadow AI? How to Find Unsanctioned AI Across Your Cloud and Company</title>
    <link>https://certisafeai.com/articles/what-is-shadow-ai/</link>
    <guid>https://certisafeai.com/articles/what-is-shadow-ai/</guid>
    <pubDate>Tue, 17 Nov 2026 12:00:00 GMT</pubDate>
    <category>AI Governance</category>
    <description>Shadow AI is AI used or deployed without security's knowledge — from employees pasting data into chatbots to forgotten cloud AI resources. Learn where it hides and how to find it on Azure, AWS and Google Cloud.</description>
  </item>
  <item>
    <title>Azure OpenAI Security Checklist: 12 Settings to Verify Before Production</title>
    <link>https://certisafeai.com/articles/azure-openai-security-checklist/</link>
    <guid>https://certisafeai.com/articles/azure-openai-security-checklist/</guid>
    <pubDate>Mon, 16 Nov 2026 12:00:00 GMT</pubDate>
    <category>Cloud Security</category>
    <description>A practical Azure OpenAI and Azure AI Foundry security checklist — disable API keys, lock down networking, enforce Prompt Shields, add an APIM gateway and turn on Defender for AI — with Azure CLI commands.</description>
  </item>
  <item>
    <title>How to Enforce Amazon Bedrock Guardrails With IAM (So No App Can Skip Them)</title>
    <link>https://certisafeai.com/articles/enforce-bedrock-guardrails-with-iam/</link>
    <guid>https://certisafeai.com/articles/enforce-bedrock-guardrails-with-iam/</guid>
    <pubDate>Fri, 13 Nov 2026 12:00:00 GMT</pubDate>
    <category>Cloud Security</category>
    <description>Bedrock Guardrails only protect requests that include them. Learn how to use the bedrock:GuardrailIdentifier IAM condition key to make guardrails mandatory, with policy examples and testing steps.</description>
  </item>
  <item>
    <title>AI Chatbot Security Checklist: 15 Controls to Check Before (and After) Launch</title>
    <link>https://certisafeai.com/articles/ai-chatbot-security-checklist/</link>
    <guid>https://certisafeai.com/articles/ai-chatbot-security-checklist/</guid>
    <pubDate>Thu, 12 Nov 2026 12:00:00 GMT</pubDate>
    <category>Checklists</category>
    <description>A practical 15-point AI chatbot security checklist covering system prompts, data access, prompt injection, output handling, agent permissions and monitoring.</description>
  </item>
  <item>
    <title>Securing Vertex AI With VPC Service Controls: A Dry-Run-First Guide</title>
    <link>https://certisafeai.com/articles/vertex-ai-vpc-service-controls/</link>
    <guid>https://certisafeai.com/articles/vertex-ai-vpc-service-controls/</guid>
    <pubDate>Fri, 06 Nov 2026 12:00:00 GMT</pubDate>
    <category>Cloud Security</category>
    <description>VPC Service Controls stop data exfiltration from Vertex AI even when credentials leak. Learn which services to include, how to roll out in dry-run mode, and the common pitfalls.</description>
  </item>
  <item>
    <title>System Prompt Leakage: What It Is and How to Prevent It</title>
    <link>https://certisafeai.com/articles/prevent-system-prompt-leakage/</link>
    <guid>https://certisafeai.com/articles/prevent-system-prompt-leakage/</guid>
    <pubDate>Tue, 03 Nov 2026 12:00:00 GMT</pubDate>
    <category>Prompt Security</category>
    <description>System prompt leakage exposes your chatbot's hidden instructions, keys and business rules. Learn how attackers extract prompts and the controls that actually prevent damage.</description>
  </item>
  <item>
    <title>What Is Prompt Injection? Direct vs. Indirect Attacks Explained</title>
    <link>https://certisafeai.com/articles/what-is-prompt-injection/</link>
    <guid>https://certisafeai.com/articles/what-is-prompt-injection/</guid>
    <pubDate>Wed, 21 Oct 2026 12:00:00 GMT</pubDate>
    <category>Prompt Security</category>
    <description>Prompt injection is the #1 LLM security risk. Learn the difference between direct and indirect prompt injection, real-world examples, and how to reduce the damage.</description>
  </item>
  <item>
    <title>How to Stop Your AI Chatbot From Leaking Customer Data</title>
    <link>https://certisafeai.com/articles/stop-ai-chatbot-leaking-customer-data/</link>
    <guid>https://certisafeai.com/articles/stop-ai-chatbot-leaking-customer-data/</guid>
    <pubDate>Wed, 07 Oct 2026 12:00:00 GMT</pubDate>
    <category>Data Protection</category>
    <description>AI chatbots can expose PII, order details and internal notes. Learn the five most common causes of chatbot data leaks and the controls that prevent them.</description>
  </item>
  <item>
    <title>Securing RAG: How to Keep Your Vector Database From Leaking Documents</title>
    <link>https://certisafeai.com/articles/securing-rag-vector-databases/</link>
    <guid>https://certisafeai.com/articles/securing-rag-vector-databases/</guid>
    <pubDate>Wed, 23 Sep 2026 12:00:00 GMT</pubDate>
    <category>RAG &amp; Agents</category>
    <description>Retrieval-augmented generation (RAG) can expose documents users were never meant to see. Learn how to enforce permissions, isolate tenants and vet content in your vector database.</description>
  </item>
</channel>
</rss>
