Akamai introduces AI-powered WAF detection architecture

Akamai has introduced an AI-powered web application firewall (WAF) detection engine designed to accelerate the mitigation of emerging threats. The architecture utilizes a decoupled approach that separates offline pattern discovery from active edge enforcement to ensure that unvalidated AI models do not interact directly with live traffic.

To maintain reliability and prevent the blocking of legitimate transactions, the system employs a six-stage safety pipeline. This process includes automated data sanitization, static backtesting, sandboxed shadow evaluations, and tiered network rollouts. The framework emphasizes deterministic governance through strict latency benchmarks, zero-tolerance false-positive testing, and real-time telemetry auto-rollbacks. Additionally, human-in-the-loop oversight is integrated to ensure all candidate signatures feature inspectable, human-readable logic.