The WAF gap no one is talking about
Your WAF is doing its job. It's blocking SQLi, XSS, and the usual suspects. But here's the problem: it wasn't built for APIs, and it definitely wasn't built for AI agents.
APIs now power nearly every digital experience. And AI agents — the automated systems that access your APIs at machine speed, at machine scale — are the fastest-growing source of that traffic. When a developer spins up an MCP server, when an internal copilot starts calling your internal APIs, or when an autonomous agent begins iterating on a task, it generates a kind of traffic that legacy WAF rules simply don't know how to interpret.
The recent Hugging Face incident made this concrete. Attackers exploited exactly the kind of gap that traditional WAF rulesets leave open: API-layer and AI-adjacent attacks that don't match classic signatures and fly under the radar of existing tools.
That gap is now closed.
Introducing Salt Managed Rules for AWS WAF
Today, we're announcing the availability of Salt Managed Rules for AWS WAF, available now in AWS Marketplace. This is the first AWS WAF Partner Managed Ruleset purpose-built to protect both APIs and the AI agents that drive them.
Existing AWS WAF customers can subscribe directly from the AWS console and attach the ruleset to their web ACLs in minutes. No new agents to deploy. No architecture changes. No long procurement cycles.
What the ruleset actually does
Advanced API threat detection
The ruleset blocks common and complex API attack vectors that standard WAF rules miss: credential brute force, excessive GraphQL queries, SSRF, prototype pollution, and JWT-based anomalies. These are the real attack patterns targeting production APIs today.
Industry-first MCP awareness
This is a first in the industry. The ruleset identifies and labels traffic from MCP endpoints, blocks unauthenticated MCP access, and gives security teams deeper observability into MCP interactions inside their AWS WAF environments. As MCP adoption accelerates, this visibility is no longer optional.
Context-aware rate limiting
Not all rate limiting is equal. The ruleset applies smart limits to sensitive parameters like user IDs and email addresses, stopping enumeration and abuse patterns that standard rate rules miss entirely.
Security signal enrichment
The ruleset labels critical request attributes including auth headers, user identifiers, and GraphQL queries. That enrichment flows into your downstream analytics and detection systems, boosting fidelity and reducing noise across your entire security stack.
Why this matters now
"AI agents are transforming how applications are built, and APIs are the layer where those agents act," said Roey Eliyahu, CEO and Co-founder of Salt Security. "By bringing Salt's API and agentic security intelligence directly into AWS WAF as managed rules, we're giving every AWS customer an easy button that deploys in minutes, so organizations can immediately see and stop the API and AI agent threats that legacy rules were never built to catch."
Salt has been building behavioral detection for APIs since 2016. That depth of expertise is what makes this ruleset different from anything else on the market. We didn't bolt AI security onto an existing product. We built detection for how agents actually behave, and we packaged it so any AWS WAF customer can get that protection today.
Get started
Salt Security Managed Rules for AWS WAF is available now across all commercial AWS Regions and globally via Amazon CloudFront.
Subscribe from the AWS Marketplace and attach the ruleset to your existing web ACLs directly from the AWS console. Learn more at the AWS Marketplace listing.
Questions? Reach out to the Salt team at salt.security.
FAQs about the Salt Security AWS WAF Managed Ruleset
Does the AWS WAF managed ruleset require any architecture changes?
No. If you already use AWS WAF, you don't need to change your architecture. You can attach the ruleset to your existing web ACLs from the AWS console. There are no proxies, no traffic redirection, and no new infrastructure to run. AWS confirms that you can subscribe to and add the rule group to a web ACL in the AWS WAF console through AWS Marketplace, with no additional configuration.
The ruleset runs inside AWS WAF itself. All partner-managed rules run on AWS WAF's managed infrastructure and protect applications on Amazon CloudFront, Application Load Balancers, or Amazon API Gateway. It is available across all commercial AWS Regions and globally through Amazon CloudFront. The only prerequisite is an AWS WAF web ACL in front of the resources you want to protect.
Want to see it in action? Watch the step-by-step deployment walkthrough to learn how to subscribe and attach the ruleset to your web ACLs.
How does it protect AI agents specifically?
AI agents reach your business through APIs and MCP servers, so the ruleset protects that layer. It does this in four ways:
- MCP awareness. The ruleset identifies traffic from AI agents and MCP endpoints, blocks unauthorized access through a dedicated unauthenticated_mcp_access rule, and labels agent-like behavior so you can observe it more closely. Salt describes this MCP awareness as an industry first.
- Context-aware rate limiting. Agents act at machine speed, so enumeration and abuse can grow quickly. The ruleset applies smart rate limits to sensitive parameters such as user_id, email, and numeric-only values to stop these patterns.
- API attack detection. Agents call APIs, so the ruleset also covers the attacks that target them. It detects credential brute force, excessive GraphQL queries, server-side request forgery (SSRF), prototype pollution, and JSON Web Token (JWT) anomalies.
- Signal enrichment. The ruleset labels request attributes, including authorization headers, user identifiers, and GraphQL queries. Your downstream analytics and detection tools can use these labels to spot agent-driven activity more accurately.
The rules also stay current without work on your side. Salt Labs security experts continuously monitor, maintain, and update the rulesets to address emerging API, AI, and agentic threats.
Are there extra AWS costs I should expect?
Yes. There are two kinds of charges.
- Salt ruleset charges (billed through AWS Marketplace). Pricing is usage-based, with a base monthly fee per region (pro-rated by the hour) plus a per-request fee per region. Subscriptions have no end date, and you can cancel at any time.
- Standard AWS WAF charges. Seller fees for a Marketplace rule group are added on top of your normal AWS WAF fees. If you already run AWS WAF, adding the Salt rule group typically adds only a small monthly rule-group charge to your AWS WAF bill. One exception: if your web ACL goes over the default 1,500 WCUs, you pay an additional per-request fee for each extra 500 WCUs.
Is there a free trial available via AWS Marketplace to test the ruleset in a staging environment?
The AWS Marketplace listing doesn't currently offer a free trial. However, because pricing is pay-as-you-go and you can cancel at any time, staging-environment testing is inexpensive. With low staging traffic, you would pay close to the $25 regional fee, pro-rated by the hour for the time the ruleset is active.
We recommend starting in Count mode. You can configure the rules to log, alert, or block, depending on your security posture and operational needs. AWS best-practice guidance suggests deploying rules in Count mode first, so they evaluate traffic without blocking it and you can observe their behavior before enforcing them. A non-production environment is a good place to start. Once you've reviewed the WAF logs and confirmed there are no false positives, switch the rules to Block.