Best-in-Class GenAI Security with CloudGuard WAF + Lakera

[Vani]

We’re excited to announce the expansion of CloudGuard WAF with integrated Lakera GenAI security now delivering prevention-first protection for Web, API, and GenAI applications. 

As enterprises rapidly adopt LLMs, generative AI, and autonomous agents, traditional security stacks can’t defend against new risks like prompt injection, data leakage, harmful content, and agent misuse. CloudGuard WAF now solves this gap without adding complexity.

What’s New

CloudGuard combines its ML-based WAF with Lakera’s GenAI engines to provide a dual-layer machine learning architecture.

Layer 1 -Supervised ML Engines

Handles 90%+ of GenAI interactions and provides prevention against known and emerging misuse patterns.

Dedicated engines secure each stage of AI interaction:

Engine	Primary Function
Prompt Injection Prevention	Blocks jailbreaks, manipulative prompts, and crafted inputs.
Data Leakage Prevention	Prevents confidential or regulated data from escaping through model responses.
Content Control	Filters harmful, policy-violating, or disallowed content.
Usage Control	Identifies abnormal access patterns, misuse, and resource abuse.

Layer 2 -Unsupervised Contextual ML

Learns the unique semantic behavior of your application, adapting continuously without manual tuning. This contextual intelligence keeps accuracy high and false positives near zero.

Refinement engines deliver contextual accuracy:

Engine	Role
User Behavior	Detects malicious deviations from normal interaction patterns.
Crowd Behavior	Learns trusted usage patterns to reduce noise and false alarms.
Trusted Users	Auto-allowlists verified inputs for frictionless experience.
Semantic Engine (Patent Pending)	Interprets natural language prompts to understand what is valid for your application—without signatures.

Why It Matters

• Real-time GenAI threat prevention
• Near-zero false positives, zero tuning
• Secures prompts, model outputs, APIs & agents
• Support 100+ native languages

This enhancement marks a major step toward our unified AI Security Platform, enabling organizations to innovate with GenAI without compromising security.

Read more- Best-in-Class GenAI Security with CloudGuard + Lakera

[the_rock]

I always truly enjoy reading these.