Skip to content
Product

Why We Submitted to NIST Twice -- And What It Means for Enterprise Agent Governance

JieGou is the only AI agent platform with dual NIST submissions on agent security and identity. Here's why we submitted, what we covered, and why it matters for enterprise procurement.

JT
JieGou Team
· · 3 min read

Two Submissions, One Mission

In 2025-2026, NIST issued two separate calls for input on AI agent governance:

  1. NIST-2025-0035: AI Agent Security — how should autonomous AI agents be secured in enterprise environments?
  2. NCCoE Agent Identity & Authorization — how should AI agents be identified, authenticated, and authorized?

We submitted detailed responses to both. Not because we had to. Because we could.

What We Submitted

Submission 1: AI Agent Security (NIST-2025-0035)

Our response addressed agent identity, authorization, tool governance, audit logging, and multi-agent security boundaries. Every recommendation was grounded in JieGou’s production infrastructure — not theoretical frameworks, but systems running in enterprise environments.

Key areas covered:

  • Agent identity and authentication (SSO/SAML/OIDC)
  • Tool governance and sandboxing (245 governed MCP servers)
  • Multi-agent security boundaries (cascading hierarchy controls)
  • Audit logging with tamper-evident records

This submission is confirmed on Regulations.gov.

Submission 2: Agent Identity & Authorization (NCCoE)

Our NCCoE response covered all 6 question categories in the concept paper, grounded in JieGou’s production agent identity infrastructure:

  • SSO/SAML/OIDC integration for agent authentication
  • 6-role RBAC with 20 granular permissions
  • Tool approval gates with configurable policies
  • Cascading agent hierarchy for multi-agent deployments
  • Shared memory isolation between agent instances
  • Comprehensive audit logging across all agent actions

Why This Matters

For Enterprise Procurement

NIST alignment is increasingly referenced in federal contracts, government-adjacent RFPs, and enterprise procurement criteria. Having verifiable NIST submissions — not just claims of alignment — provides procurement teams with evidence they can independently verify.

For the Industry

NIST standards shape how the U.S. government defines AI agent governance requirements. These standards cascade into enterprise compliance frameworks, industry certifications, and regulatory expectations. By participating in the standards-making process, we help ensure that agent governance requirements reflect production realities.

For Competitive Credibility

No other AI agent platform — not OpenAI/Frontier, not Salesforce, not ServiceNow, not Microsoft, not CrewAI, not LangGraph — has submitted to either NIST call. JieGou has submitted to both.

This isn’t a claim that can be retroactively matched. NIST RFI comment periods close permanently. The submissions are public records. This moat only strengthens with time.

The Governance Stack Behind the Submissions

Every NIST focus area maps to production governance infrastructure:

NIST Focus AreaJieGou Governance Layer
Agent IdentityLayer 1: Identity (SSO/SAML/OIDC)
AuthorizationLayer 2: RBAC (6 roles, 20 permissions)
Tool GovernanceLayer 3: Tool Approval Gates
Escalation ControlsLayer 4-5: Trust Escalation + Approval Workflows
Multi-Agent SecurityLayer 6-7: Agent Hierarchy + Memory Isolation
Audit & TraceabilityLayer 8: Audit Logging (tamper-evident)
Policy EnforcementLayer 9-10: Compliance Policies + Department Scoping
Quality MonitoringLayer 10: Quality Monitoring + GovernanceScore

What Comes Next

These submissions are the beginning, not the end. As NIST develops AI agent governance standards, JieGou will continue to participate in the standards-making process. Our goal: ensure that the standards governing AI agents reflect the depth required for enterprise deployment.


View the NIST Submissions page. Explore the 10-Layer Governance Stack. Calculate your GovernanceScore.

NIST compliance governance standards enterprise
Share this article

Enjoyed this post?

Get workflow tips, product updates, and automation guides in your inbox.

No spam. Unsubscribe anytime.