Akamai offers Firewall for AI, a purpose-built security product that inspects both the inputs sent to and outputs generated by customers' AI applications and large language models, detecting threats like prompt injection attacks, data leaks, and toxic content in real time. The product was announced in April 2025 and was in limited availability as of that date.

Firewall for AI sits between users and a customer's deployed AI application or LLM, analyzing inbound prompts for prompt injection, jailbreaks, remote code execution attempts, and unauthorized queries, and reviewing AI-generated responses for data leakage, toxicity, hallucinations, and compliance violations. It can be deployed via Akamai's edge platform or through a REST API, covering cloud, hybrid, and on-premises environments. Akamai also announced API LLM Discovery alongside it, which automatically discovers and inventories GenAI and LLM API endpoints within a customer's environment.