The definition of the "edge" is changing.
For years, security teams have focused on the traditional perimeter: web applications, public APIs, and user interfaces. We built firewalls, deployed WAFs, and established strict access controls to keep bad actors out.
But with the rapid adoption of Agentic AI, the perimeter has expanded. Today, your "edge" isn't just where users connect to your apps; it's where AI agents connect to your data.
At the center of this new architecture is the Model Context Protocol (MCP). MCP servers act as universal translators, enabling Large Language Models (LLMs) to fetch documents, query databases, and execute code. They are powerful, proliferating rapidly, and, until now, have been a massive blind spot for security teams.
The MCP Blind Spot
MCP servers serve as the critical conduit through which AI agents take action, view data, and interact with your internal and external systems. They are designed to take natural language instructions and turn them into concrete execution. While this is great for automation, it creates a high-stakes target for attackers.
Without proper oversight, MCP servers can be exploited to:
- Execute unauthorized code via prompt injection.
- Exfiltrate sensitive data by tricking an agent into reading files it shouldn't access.
- Bypass authentication controls that were designed for humans, not machines.
Most organizations don't even know how many MCP servers they have running, let alone which ones are exposed to the internet or being abused.
Extending AWS WAF to the AI Era
Today, we are thrilled to announce a new capability that closes this gap for AWS customers. Salt Security is extending its behavioral threat protection to detect and block malicious intent targeting MCP servers, directly through your existing AWS WAF.
This isn't just about adding a few static rules. It's about bringing intelligence to your edge.
By combining Salt’s deep understanding of API and AI behavior with the robust enforcement of AWS WAF, we allow you to:
- Discover the Unknown: Our MCP Finder technology maps your entire MCP footprint, external, internal, and shadow, giving you a complete inventory of your AI Action Layer.
- Detect Intent: We don't just look for bad signatures; we analyze the behavior of the agents and users interacting with your MCPs. We can spot when an agent is being manipulated or when an MCP server is being used for reconnaissance.
- Block at the Edge: When a threat is detected, Salt automatically updates your AWS WAF to block the attack in real time. This stops the threat before it ever reaches your sensitive data or applications.
Why This Matters
For security teams, speed and simplicity are everything. You don't have time to deploy and manage a separate security stack for every new AI tool your developers adopt.
With this integration, you can leverage the AWS WAF investment you’ve already made to solve a cutting-edge problem. You get immediate, automated protection against AI-specific threats without the operational overhead of new appliances or agents.
The Future is Agentic. Secure It.
Agentic AI is not a fad; it is the future of enterprise architecture. But innovation cannot come at the cost of security. By extending your perimeter protection to include MCP servers, you ensure your organization can build, deploy, and scale AI agents with confidence.
If you want to learn more about Salt and how we can help you, please contact us, schedule a demo, or visit our website. You can also get a free API Attack Surface Assessment from Salt Security's research team and learn what attackers already know.