Cloudflare AI Security

Interactive visual walkthroughs of how Cloudflare secures AI workloads — from workforce GenAI usage to protecting AI-powered applications.

1 – 4 Secure AI Usage & Applications

Use Case 1 👤 Human → Cloudflare → 🤖 AI

Secure Workforce Use of GenAI

Discover and control how employees interact with GenAI tools like ChatGPT, Claude, and Gemini through Cloudflare One SASE.

Gateway Access DLP CASB RBI AI Gateway

Explore flow → Use Case 2 🤖 Agentic AI → Cloudflare → 🗂 Resources

Govern AI Agents

Secure interactions between human users and AI agents via MCP (Model Context Protocol) with Cloudflare Access controls.

Access MCP Portals DLP for MCP Workers

Explore flow → Use Case 3 👤 Human / 🤖 Agent → Cloudflare → 🤖 AI

Build Securely with AI

Build AI applications and agents with built-in controls using AI Gateway features: caching, rate limiting, guardrails, and DLP.

AI Gateway Guardrails DLP Dynamic Routing Workers

Explore flow → Use Case 4 👤 End User → Cloudflare → 🤖 Your AI App

Protect AI-Powered Apps

Defend applications with AI endpoints against data exfiltration, abuse, prompt injection, and volumetric attacks.

WAF Bot Management DDoS AI Security for Apps API Shield Rate Limiting

Explore flow →

5 – 7 Secure AI Infrastructure & Agents

Use Case 5 🤖 AI Agent → Cloudflare → 🗂 Tools / APIs

Secure Self-Hosted AI Agents

Run AI agents on Cloudflare with defense-in-depth: isolated sandboxes, AI Gateway for LLM calls, and proxied external API access with credential injection.

Access Sandbox SDK AI Gateway Secrets Store Browser Rendering R2

Explore flow → Use Case 6 🤖 AI Agent → Cloudflare → ⚡ Sandboxed Code

Secure AI Code Execution

Safely execute AI-generated code in isolated Dynamic Worker sandboxes with millisecond startup — 100x faster than containers.

Dynamic Workers Codemode Workers RPC AI Gateway Agents SDK

Explore flow → Use Case 7 🤖 Agent → Cloudflare → 🤖 Agent

Secure AI-to-AI Communication

Orchestrate multiple AI agents with mTLS identity, per-tool authorization, durable execution via Workflows, and shared knowledge.

Agents SDK Access + mTLS MCP Portals Workflows AI Search Queues

Explore flow →