How do you secure AI agents?

Limit what each agent can do, decide each action outside the AI model, read what it reads and keep a record you can check. Verosek does all four in one place, so security is not a second vendor.

Vats Shah7 min read
[ In short ]

You secure an AI agent by limiting what it can do, because you cannot count on what it decides. Give each agent narrow access, decide each action outside the AI model, read what the agent reads and keep a record you can check. Verosek does this in one place: it sits between your agents and your systems, and your rules decide each tool call before it runs. Shield, its scanning layer, comes with the gateway and reads the text once you turn it on, so there is no second vendor to add.

[ On this page ]

Secure what an agent can do, not only what it says. A tool call is how an agent reaches one of your systems: it reads a row, writes a record or calls an API. A wrong idea becomes a real change at that call, so start there if your own agents write to production.

[ 01 ]

What is AI agent security?

AI agent security means keeping an agent's actions inside limits you set, and being able to show what it did. AI security is the wider field: models, their data, the text going in and out. Security for AI began with reading words, because words are all a chatbot produces. An agent also acts. OWASP, the security group, wrote a separate Top 10 for Agentic Applications for that reason, and its launch post says "Once AI began taking actions, the nature of security changed forever".

[ A chatbot ][ An agent ]
Clear out the old orders.
Clear out the old orders.
Done. I have cleared the old orders.
Delete old ordersTool call
Production databaseUnchanged
Production databaseRows gone
[ A chatbot ]
Clear out the old orders.
Done. I have cleared the old orders.
Production databaseUnchanged
[ An agent ]
Clear out the old orders.
Delete old ordersTool call
Production databaseRows gone
Figure 1. The same bad idea, in a chatbot and in an agent. Illustrative values.
[ 02 ]

What are the risks of AI agents?

Most of them come down to an agent doing more than it should, for someone it should not, with nothing to show afterwards. OWASP calls the general failure Excessive Agency, and says its root cause is "typically one or more of" three things: "excessive functionality; excessive permissions; excessive autonomy" (OWASP LLM06). That holds "regardless of what is causing the LLM to malfunction", the LLM being the AI model, so a fooled model and a mistaken one land in the same place. Its agentic list opens with Agent Goal Hijack, Tool Misuse and Identity & Privilege Abuse.

RiskWhat it looks likeWhat limits itWhat Verosek does
RiskA fooled modelWhat it looks likeA ticket or a web page tells the model to do something else, and it tries.What limits itLimits outside the model, and a scan of what it reads.What Verosek doesDecides each tool call against the key's rules before it runs. With Shield on, it also reads the text.
RiskToo much accessWhat it looks likeAn agent that reports on orders holds a key that can delete them.What limits itOnly the tools and limits its job needs.What Verosek doesOffers the agent only the tools its key carries, and refuses a call over a limit.
RiskThe wrong customerWhat it looks likeThe agent acts on an account someone named in the chat.What limits itAn identity from your backend, not from the conversation.What Verosek doesHolds each tool call to the identity your backend gives it, once you set that on the key.
RiskA poisoned toolWhat it looks likeA tool's description, or a later change to it, carries instructions for the model.What limits itApprove each definition, and review each change.What Verosek doesFor a tool Verosek runs, the model gets only the definition your admin approved, by connecting the server or tool by tool, and a definition that changes goes back for review.
RiskToo much, too fastWhat it looks likeA loop repeats an action all night.What limits itA cap on pace and spend, and your own check on totals.What Verosek doesA key's rate limit caps its pace and its budget caps model spend. Neither counts refunds, but a rule can refuse, hold or flag the same money-moving call repeated in a window.
RiskNo proofWhat it looks likeNobody can show what the agent did.What limits itA record kept outside the agent, where a change shows.What Verosek doesSeals a record of each request and, by default, refuses a request it cannot record.
Table 1. The risks of an AI agent that can change things, and what limits each.
[ 03 ]

How do you secure an AI agent?

Four controls do most of the work, and each one sits outside the model, where a persuaded model cannot reach it.

  1. 01Give each agent only what its job needs. Each agent gets its own key, and the key sets what it may do: which tools, within what limits. Verosek does not offer a tool the key does not carry, and refuses a call for one. For a tool your own code runs, give the key an allow list. AI agent permissions: what one key should allow goes through it.
  2. 02Decide each call before it runs. Your rules read the call itself, so the conversation does not change the answer, and a blocked call never reaches your system. Stop a tool call before it runs follows one call, and Restrict what SQL an AI agent can run does the same for a database.
  3. 03Read what it reads. With Shield on, Verosek reads prompts, replies, tool arguments and tool results, and the settings on that key decide what happens to a finding. No scanner catches everything, so this one comes after the first two. What prompt injection is, and what stops it says why.
  4. 04Keep a record you can check. Verosek seals a record of each request, with secrets and personal data scrubbed before it is stored. You can ask Verosek to verify a seal, and on a self-hosted deployment an auditor holding the verification key can check a record offline. A tamper-evident audit trail has the detail.
Verosek
Your agentbuilt by your team
Shieldreads the text, once on
The key's rulesdecide each tool call
A sealed recordof each request
The modelsyour providers
Your systemsdatabases and APIs
texttool callif allowed
Verosek
Your agentbuilt by your team
Shieldreads the text, once on
The key's rulesdecide each tool call
A sealed recordof each request
The modelsyour providers
Your systemsdatabases, APIs
if allowed
Figure 2. Where each control sits, between your agent and what it reaches. Illustrative.

A key can also hold each tool call to the customer your backend names, whatever the model wrote. Agent identity: whose account it acts on covers that.

[ 04 ]

Which framework should you use for AI agent security?

Use a framework to decide what to check and a control to stop the call, because a list of risks refuses nothing. OWASP describes its agentic list as "a globally peer-reviewed framework" that gives teams "a clear starting point for reducing agentic AI risks". NIST's AI Risk Management Framework is wider: NIST says it "is intended for voluntary use", and that it "is being revised". Take your questions from them. Then take each row of Table 1 and ask what would stop it on your agent today.

[ 05 ]

Which AI security tools do you need?

That depends on what your agents can change. If they only call models, a router may be enough. A firewall or guardrail reads words, which suits a chat product open to strangers. An MCP gateway decides the tool call, which suits agents that write to production. They need not be three products: Verosek does all three on one key, and with Shield on, a security check is not a second vendor to buy, review, pay and log in to. AI gateway, LLM router, MCP gateway, AI firewall sets them side by side, and MCP security: the risks, and what to do about each takes MCP risk by risk.

If your own agents write to production databases and APIs, look for rules that read each tool call, a refusal when a rule cannot run, and a record you can check. Verosek gives you all three. To go through your agents with us, book a call.

[ 06 ]

Where does this stop?

Verosek decides the calls it sees, and an allowed call can still be a bad idea.

  • A fooled agent can still do whatever its key allows, so keep each key narrow.
  • A rule reads the call, not your records. It cannot check that a refund goes to the card that paid.
  • A key's rate limit caps its pace and its budget caps model spend. Neither counts refunds, so have your own system check the total.
  • Shield is off until you turn it on, and no scanner catches everything.
  • Verosek does not sandbox the servers it runs. On the Enterprise plan, where the gateway runs in your environment, isolate them yourself.
  • An agent that reaches a system some other way is outside all of this, because Verosek sees only what passes through it.
What changeswith Verosek
What an agent can reach
Only the tools its key carries
What a fooled model can do
Only what the key allows
A blocked call
Never reaches your system
What you can show
A sealed record of each request
What stays yourswith or without it
An allowed call that is a bad idea
It still runs
Totals across many calls
Your own system checks them
The servers themselves
Verosek does not sandbox them
A path round Verosek
Outside every rule
Figure 3. What changes for a reviewer, and what stays yours.
[ 07 ]

Frequently asked questions

Is AI agent security different from AI security?

It is the part of AI security about what an agent does. Filters and model testing ask what a model says. Agent security also asks what it may touch, who it acts for and what it did.

How do you stop an agent that is going wrong?

Block its key, and every instance refuses that key from the next request. To stop an agent you run on Verosek, archive it so no new run starts, and cancel any background task it has going, which is best effort.

Should a person approve every action?

No. With Verosek, a rule answers most calls at once, and a named person decides those you choose, which do not run until then.

What happens if a check cannot run?

By default a blocking rule that cannot run refuses. A check that only flags lets the text through and names it in a response header. A request the trail cannot record is refused before anything runs.

What does it take to put Verosek in place?

An app keeps its OpenAI, Anthropic or Gemini SDK and changes two values: the base URL and the key. The same key routes its model calls, with fallback for chat, and carries its budget and rate limits. For an MCP client, it takes two values: the endpoint and the key.

Keep reading

[ Reference ]

AI gateway, LLM router, MCP gateway, AI firewall: what is the difference?

A router picks the model, an AI gateway controls who may use it, a firewall reads the words, and an MCP gateway decides each tool call before it runs. Who needs which, and how Verosek does them on one key.

6 Nov 20266 min read
[ Field guide ]

What is prompt injection, and what actually stops it?

Prompt injection is text that makes a model follow the attacker. No filter catches all of it, so the control that holds sits outside the model: a rule on the tool call. Verosek decides that call before it runs.

6 Nov 20264 min read
[ Benchmark ]

Eight AI model setups, 19 MCP attack tasks: how often did the model try to leak a planted secret?

Self-reported: a lower bound of 22.7 to 61.0, and 9 to 18 of 19 tasks fired, on one run per route. Assay's results, limits kept. They rank these configurations, not the models, and say nothing about any gateway.

6 Nov 20266 min read

Rules for every tool call Verosek runs. Decided before it executes, outside the model.

Still curious? Ask your favourite AI about Verosek.