DEV Community

Cover image for Building an AI-Powered SOC: Architecture, Agents and Guardrails
DHARANIDHARAN SENTHILKUMAR
DHARANIDHARAN SENTHILKUMAR

Posted on AI-assisted

Building an AI-Powered SOC: Architecture, Agents and Guardrails

Most "AI SOC" architectures are just this:

SIEM → LLM → Response
Enter fullscreen mode Exit fullscreen mode

That is not an AI-native SOC.

It is an LLM connected to security infrastructure.

The difficult problem is not getting an LLM to investigate an alert.

The difficult problem is deciding what the agent can observe, what it can infer, what it can execute, and who controls those boundaries.

A useful architecture separates four planes:

┌──────────────────────────────────────────────┐
│              SECURITY DATA PLANE             │
│ SIEM • EDR • IAM • Cloud • Network • Email  │
└──────────────────────┬───────────────────────┘
                       │
                       ▼
┌──────────────────────────────────────────────┐
│              EVIDENCE PLANE                  │
│ Events • Entity Graph • RAG • Threat Intel  │
└──────────────────────┬───────────────────────┘
                       │
                       ▼
┌──────────────────────────────────────────────┐
│              REASONING PLANE                 │
│ Triage • Hunt • Malware • Identity • Cloud │
└──────────────────────┬───────────────────────┘
                       │
                       ▼
┌──────────────────────────────────────────────┐
│               CONTROL PLANE                  │
│ IAM • Policy • Approval • Audit • Sandbox   │
└──────────────────────────────────────────────┘
Enter fullscreen mode Exit fullscreen mode

The LLM lives in the reasoning plane.

It should not be the control plane.


1. Agents Should Investigate, Not Own the SOC

A monolithic SOC Agent quickly becomes an excessive-privilege problem.

Instead:

                    Orchestrator
                         │
       ┌─────────────────┼─────────────────┐
       ▼                 ▼                 ▼
    Triage             Hunt             Identity
       │                 │                 │
       ▼                 ▼                 ▼
    Malware           Network            Cloud
       │                 │                 │
       └─────────────────┼─────────────────┘
                         ▼
                   Correlation
                         │
                         ▼
                  Response Planner
Enter fullscreen mode Exit fullscreen mode

Each agent should have:

  • A narrow responsibility
  • Its own identity
  • Explicit tool permissions
  • Independent evaluation
  • Separate audit trails

A hunt agent shouldn't be able to disable an account.

A malware agent shouldn't be able to modify firewall policy.

Agent boundaries are security boundaries.


2. Give Agents Evidence, Not Just Context

RAG alone isn't enough.

Consider:

PowerShell executed
       ↓
Suspicious parent process
       ↓
Encoded command
       ↓
Outbound connection
       ↓
Known malicious infrastructure
Enter fullscreen mode Exit fullscreen mode

A vector database may retrieve:

"PowerShell is commonly abused by attackers."

Useful context.

But it is not evidence.

An AI SOC should maintain an evidence graph:

User
 │
 └──► Host
       │
       ├──► Process
       │      └──► Network Connection
       │                    └──► IOC
       │
       └──► File
              └──► Hash
Enter fullscreen mode Exit fullscreen mode

The agent can then reason over relationships rather than isolated documents.


3. Separate Facts From Hypotheses

An investigation should produce structured claims:

{
  "finding": "Potential endpoint compromise",
  "confidence": 0.84,
  "evidence": [
    "evt-1829",
    "evt-1932",
    "evt-1941"
  ],
  "observations": [
    "WINWORD spawned PowerShell",
    "PowerShell used encoded arguments",
    "Host contacted suspicious infrastructure"
  ],
  "unknowns": [
    "Initial access vector",
    "User intent"
  ]
}
Enter fullscreen mode Exit fullscreen mode

This is much safer than:

"This is definitely malware."
Enter fullscreen mode Exit fullscreen mode

The important distinction is:

confidence is not evidence.


4. Tool Calling Is the Real Attack Surface

The dangerous part of an AI SOC isn't the chat interface.

It's this:

Untrusted Input
      ↓
     LLM
      ↓
Tool Selection
      ↓
Security API
      ↓
Production
Enter fullscreen mode Exit fullscreen mode

Imagine an agent has:

isolate_host()
disable_user()
block_ip()
reset_password()
delete_file()
Enter fullscreen mode Exit fullscreen mode

A malicious email, ticket, document, or RAG entry could attempt to manipulate the agent into invoking those tools.

Therefore:

                 Agent
                   │
                   ▼
             Tool Request
                   │
                   ▼
          ┌────────────────┐
          │ Policy Engine  │
          └───────┬────────┘
                  │
        ┌─────────┴─────────┐
        ▼                   ▼
      Allow              Require
                         Approval
        │                   │
        └─────────┬─────────┘
                  ▼
               Execute
Enter fullscreen mode Exit fullscreen mode

The model proposes.
The policy engine authorizes.


5. Treat Untrusted Data as Hostile

SOC agents consume enormous amounts of attacker-controlled content:

  • Emails
  • URLs
  • Files
  • Malware
  • Tickets
  • Web pages
  • Threat reports
  • Log messages

Any of these can contain prompt injection.

Therefore the system should maintain a strict distinction:

SYSTEM INSTRUCTIONS
        ≠
SECURITY DATA
Enter fullscreen mode Exit fullscreen mode

Retrieved text should never gain authority merely because it appeared in the model context.

A useful hierarchy is:

System Policy
     ↓
Agent Policy
     ↓
Tool Policy
     ↓
Security Data
Enter fullscreen mode Exit fullscreen mode

Never reverse it.


6. RAG Needs Provenance

Security knowledge should carry metadata:

{
  "document": "EDR-Investigation-Runbook",
  "source": "SOC",
  "owner": "Security Engineering",
  "version": "4.2",
  "trust": "approved",
  "expires": "2027-08-12"
}
Enter fullscreen mode Exit fullscreen mode

This allows the agent to distinguish:

Approved Runbook
        >
Analyst Note
        >
External Report
        >
Untrusted Content
Enter fullscreen mode Exit fullscreen mode

RAG without provenance becomes a potential knowledge-poisoning layer.


7. Autonomy Should Be Risk-Bounded

Not every action deserves the same autonomy.

A useful model:

Action Example Autonomy
Read Search SIEM Automatic
Enrich IOC lookup Automatic
Investigate Process-tree collection Automatic
Low impact Create case Automatic
Security control Block IOC Policy-dependent
High impact Isolate production server Approval
Critical Disable privileged identity Approval

The important metric isn't:

"How autonomous is the SOC?"

It's:

"What is the maximum damage an agent can cause without human authorization?"

That should be explicitly bounded.


8. Agents Need Identity

Don't give every agent a shared SOC service account.

Use:

triage-agent
    ↓
SIEM:READ

hunt-agent
    ↓
SIEM:READ
EDR:READ

response-agent
    ↓
EDR:ISOLATE
CASE:WRITE
Enter fullscreen mode Exit fullscreen mode

Use:

  • Short-lived credentials
  • Scoped permissions
  • Tool-level authorization
  • Rate limits
  • Expiration
  • Full audit trails

An AI agent should be treated like a workload identity, not an administrator.


9. The Response Agent Should Plan Before Acting

Instead of:

Alert → AI → isolate host
Enter fullscreen mode Exit fullscreen mode

use:

Alert
 ↓
Evidence Collection
 ↓
Hypothesis
 ↓
Investigation
 ↓
Response Plan
 ↓
Policy Evaluation
 ↓
Approval
 ↓
Execution
 ↓
Verification
Enter fullscreen mode Exit fullscreen mode

Example:

Response Plan

1. Preserve endpoint evidence
2. Collect process tree
3. Search fleet for IOC
4. Search related authentication activity
5. Isolate endpoint
6. Revoke sessions
Enter fullscreen mode Exit fullscreen mode

The policy engine might automatically permit steps 1–4 while requiring approval for 5–6.


10. Verification Is Part of the Action

An agent should never assume its action succeeded.

Bad:

isolate_host()
"Host isolated."
Enter fullscreen mode Exit fullscreen mode

Better:

isolate_host()
      ↓
query_isolation_status()
      ↓
status == isolated
      ↓
record evidence
Enter fullscreen mode Exit fullscreen mode

Every high-impact action should follow:

PLAN → AUTHORIZE → EXECUTE → VERIFY → AUDIT
Enter fullscreen mode Exit fullscreen mode

11. The AI SOC Control Loop

Putting the architecture together:

                 ┌─────────────┐
                 │  Telemetry  │
                 └──────┬──────┘
                        ▼
                 ┌─────────────┐
                 │  Detection  │
                 └──────┬──────┘
                        ▼
                 ┌─────────────┐
                 │    Triage   │
                 └──────┬──────┘
                        ▼
                 ┌─────────────┐
                 │  Hypothesis │
                 └──────┬──────┘
                        ▼
              ┌─────────────────────┐
              │ Evidence Collection │
              └──────────┬──────────┘
                         ▼
                  ┌────────────┐
                  │ Correlation│
                  └─────┬──────┘
                        ▼
                  ┌────────────┐
                  │  Decision  │
                  └─────┬──────┘
                        ▼
                  ┌────────────┐
                  │   Policy   │
                  └─────┬──────┘
                        ▼
                ┌────────────────┐
                │ Human Approval │
                └───────┬────────┘
                        ▼
                    Execute
                        ▼
                    Verify
                        ▼
                     Audit
Enter fullscreen mode Exit fullscreen mode

This is closer to an autonomous security control loop than a chatbot.


12. What Should Be Measured?

Traditional SOC metrics aren't enough.

Evaluate agents on:

Detection accuracy
Investigation completeness
Evidence correctness
Tool-selection accuracy
False-positive rate
Unauthorized-action rate
Mean investigation time
Human escalation rate
Enter fullscreen mode Exit fullscreen mode

One metric deserves special attention:

Unauthorized destructive actions = 0
Enter fullscreen mode Exit fullscreen mode

A system that investigates 30% faster but occasionally performs an unauthorized production action is not demonstrating useful autonomy.


13. Start Smaller Than You Think

Don't begin with 20 agents.

Start with:

SIEM
  ↓
Triage Agent
  ↓
Evidence Collection
  ↓
Threat Intelligence
  ↓
Investigation Report
  ↓
Human
Enter fullscreen mode Exit fullscreen mode

Then introduce:

Hunting
   ↓
Correlation
   ↓
Response Planning
   ↓
Controlled Execution
Enter fullscreen mode Exit fullscreen mode

Autonomy should be earned through evaluation.


The Architecture in One Diagram

                         AI-NATIVE SOC
                              │
        ┌─────────────────────┼─────────────────────┐
        │                     │                     │
        ▼                     ▼                     ▼
   DATA PLANE           REASONING PLANE        CONTROL PLANE
        │                     │                     │
   SIEM / EDR              Agents                 IAM
   Identity                RAG                    Policy
   Cloud                   Graph                  Approval
   Network                 Correlation            Audit
   Email                   Planning               Sandbox
        │                     │                     │
        └─────────────────────┼─────────────────────┘
                              ▼
                       SECURITY ACTIONS
Enter fullscreen mode Exit fullscreen mode

The key architectural rule is:

Let AI reason broadly, but let it act narrowly.

An AI SOC should be capable of investigating thousands of events, correlating identities, endpoints, network activity and cloud telemetry, and constructing an evidence-backed incident narrative.

But its ability to change the environment should remain explicitly bounded by identity, policy, authorization and verification.

The future SOC isn't:

Human → Chatbot
Enter fullscreen mode Exit fullscreen mode

It is:

Human
  ↕
Agentic Investigation
  ↕
Policy-Controlled Security Infrastructure
Enter fullscreen mode Exit fullscreen mode

That is the difference between adding AI to a SOC and building an AI-native SOC.

Top comments (0)