Cloudflare AI Security for Apps Is Now Generally Available

Cloudflare

Mar 15, 2026 · Updated Apr 25, 2026

Cloudflare AI Security for Apps is now generally available, providing a security layer that discovers LLM-powered endpoints, detects prompt injection and PII leakage, and mitigates threats through WAF rules. AI endpoint discovery is now free for all Cloudflare plans.

Cloudflare launched AI Security for Apps into general availability, a security layer for AI-powered applications that covers discovery, detection, and mitigation. It sits in Cloudflare's reverse proxy in front of any app — regardless of model or provider — identifying LLM endpoints by behavior rather than URL patterns, scanning prompts always-on for injection attempts, PII extraction, and toxic topics, then letting teams act through the same WAF rules engine they already use.

New in GA: a custom topics feature lets you define your own off-limits categories and receive a relevance score to log, block, or route as you decide. Mitigation combines AI-specific signals with Cloudflare's full request context — IP reputation, browser fingerprints, botnet data — connections point solutions can't make.

Enterprise customers get full detection and mitigation now. Free, Pro, and Business plan users can start with AI endpoint discovery today through Security → Web Assets in the dashboard.

View the full update on blog.cloudflare.com

Cloudflare

@CloudflareMar 11

Cloudflare’s AI Security for Apps detects and mitigates threats to AI-powered applications. Starting today, it is generally available. https://t.co/7zqsJxUPBd

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

See all AI news & updates from Cloudflare →

Keep reading

Cloudflare Launches Dynamic Workers for Sandboxing AI-Generated Code

Cloudflare's Dynamic Workers, now in open beta for paid Workers users, runs AI-generated code in V8 isolate sandboxes that start in milliseconds — roughly 100x faster than containers. This makes per-request, per-user agent sandboxing practical at scale without concurrency limits.

Cloudflare adds xAI Grok models to AI Gateway with unified billing

Cloudflare DevelopersJun 4

Cloudflare adds xAI Grok models to AI Gateway with unified billing

Cloudflare has integrated xAI's full Grok model suite into its AI Gateway platform. This move allows developers to deploy frontier reasoning and generative media models through a single control plane, eliminating the need for separate API management and fragmented billing across providers.

Vercel Adds Natural Language Interface to Firewall for Instant Security Rules

VercelMay 15

Vercel Adds Natural Language Interface to Firewall for Instant Security Rules

Vercel launched a natural language interface for its Web Application Firewall that allows developers to generate custom security rules from plain-English descriptions. This update removes the syntax barrier for complex traffic management, enabling teams to deploy rate limits and IP blocks via the dashboard or CLI without manual configuration.