AWS adds WAF protection for Amazon Bedrock AgentCore Gateway

AWS made AWS WAF protection generally available for Amazon Bedrock AgentCore Gateway, letting security and platform teams apply consistent, customizable web protections to agentic AI workloads as they move from prototype to production. The integration guards against common web exploits and abuse patterns at the gateway layer where agents expose tools and endpoints.

Mechanically, it brings AWS's mature web-application-firewall tooling — rate limiting, managed rule sets, geo and IP controls — to the new attack surface that agents create. As agents invoke external tools and APIs, they introduce injection, abuse, and exfiltration risks that traditional app security wasn't designed for.

The launch complements AWS's Summit slate (AgentCore GA, Managed Knowledge Bases) by addressing the security half of productionizing agents — a recurring enterprise blocker. It also aligns with the week's broader security theme, including Apple shipping patches early and AWS's June Threat Technique Catalog update.

For enterprises, native WAF support reduces the bespoke security engineering needed to safely deploy agents, which AWS is betting accelerates adoption. Skeptics note that web-layer protections only cover part of agent risk — prompt injection, tool misuse, and data isolation need additional controls. What to watch: how AgentCore's security stack evolves and whether it becomes a differentiator against rival agent platforms.