Cloudflare AI Security for Apps

Cloudflare AI Security for Apps adds model-independent LLM threat detection and blocking controls to its web application firewall.

Verified 11 Nov 2026

Facts

Source per value

Runtime guardrails, as documented What the vendor documents for this capability.: Detects prompt injection and unsafe or custom topics in incoming JSON prompts; WAF rules apply blocking or challenges. 1
Data protection for AI, as documented What the vendor documents for this capability.: Detects PII in incoming prompts including contact, financial, and identity information. 1
Integrations and platforms Named clouds, models, development tools, protocols or security tools the product supports.: Requires an enabled Cloudflare WAF zone; supports JSON request bodies. 1
Licensing Open-source status or paid packaging, where documented.: LLM endpoint discovery is on all plans; AI detection fields require an Enterprise paid add-on. 1

Akamai Firewall for AI: Akamai Firewall for AI is an LLM firewall that inspects prompts and model responses at the Akamai edge or through an API.
Alice (ActiveFence): Alice provides AI safety and security for model labs, enterprises, and platforms through adversarial testing, custom guardrails, and threat intelligence.
Amazon Bedrock Guardrails: Amazon Bedrock Guardrails provides configurable safeguards for enterprise generative AI applications and agent workflows, including models hosted outside Bedrock.
Aurascape: Aurascape secures enterprise AI usage and AI agents with discovery, data protection, adversarial testing, and runtime controls.

Tell us what you need once. We send it to matching vendors and they reply to you directly.