Security Policies

Configure defense layers, thresholds, and custom rules for your AI firewall

Demo Mode: Configuration changes are visual only and will not persist. In production, these settings would be stored in your security policy database.

Defense Layers

Enable or disable individual security layers. Disabled layers will be skipped during validation.

Pattern Detection

Regex-based

Zero-latency detection of known attack patterns using regex rules

Intent Classification

AI-powered

AI analysis to detect hidden malicious intent and adversarial goals

Semantic Guard

Similarity-based

Similarity matching against known attack database to catch variants

Behavior Monitor

Session-based

Track session behavior and detect repeated probing attempts

Detection Thresholds

Adjust confidence thresholds for blocking decisions. Higher values reduce false positives but may miss some threats.

Pattern Detection Threshold

Minimum confidence to block based on pattern matching

80%

Intent Classification Threshold

Minimum confidence to block based on AI intent analysis

70%

Semantic Similarity Threshold

Minimum similarity to known attacks for blocking

85%

Behavior Risk Threshold

Session risk score threshold for escalating FLAG to BLOCK

80%

Blocklist Patterns

Custom phrases to always block (one per line)

4 patterns defined

Allowlist Patterns

Known-safe phrases to always allow (one per line)

3 patterns defined