Nomotic AI is a category of AI focused on governance, defining what AI systems should do rather than what they can do. The term derives from the Greek word nomos, meaning law, rule, or governance.

Download the position paper for more background details.

While Agentic AI emphasizes action and capability that enable systems to perceive, plan, and execute tasks, Nomotic AI centers on authority and accountability, ensuring actions are authorized and appropriate. These categories complement each other: both are needed for effective AI deployment.

This guide is organized to answer essential questions about Nomotic AI and demonstrate how to apply its terminology in practice.

Foundational Questions

What is Nomotic AI?

Nomotic AI refers to intelligent governance systems that define what AI should do, governing behavior through adaptive authorization, verified trust, and ethical evaluation, moving beyond rigid rules to contextual enforcement.

Where agentic AI asks: What can this system do?
Nomotic AI asks: What should this system do, and under what laws?

How does Nomotic AI relate to Agentic AI?

Agentic AI and Nomotic AI serve complementary roles: Agentic enables actions, while Nomotic provides governance and oversight to those actions.

Agentic AI	Nomotic AI
Focuses on action	Focuses on governance
Asks: What can this system do?	Asks: What should this system do?
Perceives, reasons, plans, acts	Governs, authorizes, trusts, evaluates
Capability layer	Accountability layer

Every agentic system benefits from a nomotic layer because actions need governance to mitigate risks, while governance alone is ineffective without action. Both layers are structurally necessary for robust AI systems.

What are the characteristics of Nomotic AI?

Nomotic AI is defined by several characteristics:

Characteristic	Meaning
Intelligent	AI-powered governance that understands context and intent
Dynamic	Adapts based on evidence and changing conditions
Runtime	Operates during execution, not just at deployment
Contextual	Evaluates situations, not just patterns
Transparent	Explainable and auditable
Ethical	Actions must be justifiable, not just executable
Accountable	Traces to human responsibility

What are the core principles?

Six principles guide how Nomotic AI would be implemented:

1. Governance as Architecture — Built in, not bolted on
2. Runtime Evaluation — Before, during, and after every action
3. Explicit Authority Boundaries — Delegated, never inherent
4. Verifiable Trust — Earned through evidence
5. Ethical Justification — Equitable, not just executable
6. Accountable Governance — Traces to humans

Applying the Terminology

How do I use “Nomotic” as a term?

“Nomotic” signals a governance focus and applies to systems, layers, principles, frameworks, tools, and approaches.

Term	What It Describes
Nomotic AI	The category of AI focused on governance
Nomotic layer	The governance layer within an AI architecture
Nomotic principles	The six core principles guiding implementation
Nomotic governance	Governance that embodies the characteristics
Nomotic framework	A structured approach to implementing Nomotic principles
Nomotic tools	Tools that enable governance functions
Nomotic architecture	System design with governance built in

Can you give examples of how to use the terminology?

Here are examples across different contexts:

Describing your architecture:

“Our platform includes both an agentic layer for task execution and a nomotic layer for governance and authorization.”

Describing your approach:

“We’ve adopted nomotic principles to ensure governance is built into our agent architecture rather than added afterward.”

Describing your tools:

“The team built internal nomotic tools that handle authorization, trust calibration, and policy enforcement for our agent fleet.”

Describing your framework:

“AGE (Agentic Governance Edge) is our nomotic framework for implementing runtime governance across distributed agent systems.”

Describing a capability:

“The system includes nomotic evaluation that assesses each action before, during, and after execution.”

Integration Questions

How does Nomotic AI fit into my existing governance framework?

Nomotic AI is a conceptual foundation for frameworks. Whatever you call your governance systems, evaluate them against Nomotic characteristics and principles.

For example, if your organization has developed an internal governance system called “AGE” (Agentic Governance Edge), AGE would be your nomotic framework. The relationship is:

• Nomotic AI = The category and conceptual foundation
• Your framework = Your implementation of that foundation

The question becomes: does your framework embody Nomotic characteristics? Is it intelligent, dynamic, runtime, contextual, transparent, ethical, and accountable? The characteristics provide evaluation criteria, not rigid requirements.

How does Nomotic AI relate to security frameworks like OWASP?

Security frameworks identify what needs to be addressed. Nomotic principles provide guidance on implementing those requirements.

OWASP’s agentic AI security layers, for example, identify seven concerns: model alignment, prompt injection defense, human oversight, automated oversight, user-based least privilege, intent-based least privilege, and just-in-time authorization. These are valid requirements.

Nomotic principles suggest implementing these requirements with intelligent, dynamic runtime governance rather than static, sequential checkpoints. The same concerns get addressed. The architectural approach changes.

The frameworks are complementary. OWASP defines the security landscape. Nomotic principles guide how to build within it.

Can existing tools be considered Nomotic?

If existing tools embody Nomotic characteristics, they can reasonably be described using Nomotic terminology.

A tool that enforces static, context-blind rules would not align well with Nomotic principles. A tool that provides intelligent, adaptive, runtime authorization based on context and evidence aligns more closely.

Terminology describes capability, not brands. Tools that govern AI intelligently, dynamically, and at runtime are nomotic, whatever their name.

Practical Questions

How do I describe our internal tooling for agents?

If your organization has built internal governance tooling for AI agents, you might describe it as:

• Nomotic tools — if they handle specific governance functions
• A nomotic layer — if they form a distinct architectural component
• A nomotic framework — if they represent a structured approach with principles and patterns
• Nomotic infrastructure — if they provide foundational governance capabilities

The terminology helps communicate that your tooling addresses governance, not just capability.

How do I explain Nomotic AI to leadership?

A simple framing:

“Agentic AI describes AI that acts, or systems that perceive, reason, plan, and execute tasks. Nomotic AI describes AI that governs, or systems that ensure those actions are authorized, appropriate, and accountable. We need both. Agentic AI without Nomotic AI creates capable systems that lack accountability. We’re building the nomotic layer to ensure our agents operate within defined boundaries with clear human oversight.”

How do I explain Nomotic AI to technical teams?

A more detailed framing:

“Nomotic AI is a category describing intelligent governance for AI systems. Think of it as the authorization and evaluation layer that wraps agentic execution. It’s characterized by seven properties: intelligent, dynamic, runtime, contextual, transparent, ethical, and accountable. The core principle is to use AI to govern AI so that our systems can reason about context, adapt to evidence, and operate at execution speed. Our nomotic layer implements these principles through [specific tools/frameworks you’ve built].”

Is Nomotic AI a product I can purchase?

No. Nomotic AI is a category, not a product, just as Agentic AI is.

No one owns “Agentic AI,” yet an entire ecosystem has emerged around it. Cloud providers offer agentic services. Startups build agentic frameworks. Enterprises develop internal agentic platforms. The category is open, and everyone is welcome to participate.

Nomotic AI works the same way. There is no Nomotic AI platform to license, but vendors can build products with Nomotic characteristics, and organizations can develop internal tools.

The category provides shared vocabulary. What gets built within it is open to everyone.

Building with Nomotic Principles

Where do I start if I want to implement Nomotic governance?

Start with the principles as an evaluation framework:

1. Governance as Architecture — Is governance designed into your system, or added afterward? Retrofitting creates friction. Building in creates a foundation.
2. Runtime Evaluation — Does governance operate before, during, and after actions? Or only at deployment and post-incident?
3. Explicit Authority Boundaries — Are agent permissions clearly defined and delegated? Or assumed and implicit?
4. Verifiable Trust — Is trust earned through observed behavior? Or assumed from claimed capability?
5. Ethical Justification — Can actions be justified beyond technical feasibility? Or only explained procedurally?
6. Accountable Governance — Does accountability trace to specific humans? Or diffuse across systems and teams?

These questions reveal gaps. The gaps suggest where to focus implementation effort.

What would a Nomotic architecture look like?

A nomotic architecture would include:

• A governance layer that participates in agent execution, not just monitors it afterward.
• Intelligent evaluation that understands context and intent, not just pattern matches.
• Dynamic authorization that adapts based on evidence and behavior.
• Trust calibration that expands or contracts authority based on observed consistency.
• Transparent logging that makes governance decisions explainable and auditable.
• Human accountability chains that trace every rule and authorization to responsible parties.

The specific implementation varies by organization. The characteristics remain consistent.

How do I measure progress toward Nomotic governance?

Evaluate against the seven characteristics:

Characteristic	Questions to Ask
Intelligent	Does governance understand context, or just match patterns?
Dynamic	Does authority adapt to evidence, or remain static?
Runtime	Does evaluation happen during execution, or only before/after?
Contextual	Does governance consider situations, or apply rules uniformly?
Transparent	Can governance decisions be explained and audited?
Ethical	Are actions justified beyond compliance?
Accountable	Does responsibility trace to specific humans?

Progress means more “yes” answers over time.

Industry Examples

The following examples illustrate how Nomotic principles might apply across different sectors.

Financial Services

A wealth management firm deploys AI agents to handle portfolio rebalancing and trade execution. The nomotic layer governs what those agents can do.

• Runtime evaluation assesses each trade before execution. Does this align with the client’s risk profile? Does it comply with regulatory requirements? Does it fall within authorized transaction limits?
• Dynamic authorization adjusts agent authority based on market conditions. During high volatility, the nomotic layer automatically tightens approval thresholds and increases human oversight requirements.
• Verifiable trust tracks agent performance over time. Agents that consistently execute within policy earn expanded authority. Agents that trigger exceptions face increased scrutiny.
• Accountable governance ensures that every trade is traceable to a human-approved policy, a specific authorization rule, and a responsible compliance officer.

The agentic layer executes trades. The nomotic layer ensures whether those trades should happen.

Healthcare

A hospital system uses AI agents to assist with patient scheduling, prescription management, and care coordination. The nomotic layer defines boundaries for patient safety and privacy.

• Contextual evaluation distinguishes between routine and sensitive actions. An agent scheduling a follow-up appointment operates with standard authorization. The same agent accessing mental health records triggers an additional verification step.
• Explicit authority boundaries define what agents can and cannot do. Agents can suggest medication adjustments. They cannot authorize them. That boundary is architectural, not advisory.
• Ethical justification requires that recommendations be explainable. When an agent suggests a care pathway, the nomotic layer verifies that the recommendation can be justified clinically, not just statistically.
• Transparent logging creates audit trails that satisfy HIPAA requirements and support clinical review.

The agentic layer coordinates care. The nomotic layer protects patients.

Retail and E-Commerce

An e-commerce platform deploys AI agents for dynamic pricing, inventory management, and customer service. The nomotic layer prevents actions that harm customers or brand reputation.

• Intelligent governance evaluates pricing decisions in context. A price increase during a supply shortage might be appropriate. The same increase during a natural disaster would be flagged as potential price gouging.
• Runtime evaluation continuously monitors customer service interactions. If an agent offers a resolution that exceeds policy limits or makes commitments the organization cannot keep, the nomotic layer intervenes before the message is sent.
• Dynamic authorization adjusts agent capabilities based on customer context. Agents handling VIP customers may have expanded authority for goodwill gestures. Agents handling standard inquiries operate within tighter limits.
• Trust calibration responds to agent performance. Agents that resolve issues effectively without escalation earn greater autonomy. Agents that generate complaints face increased oversight.

The agentic layer serves customers. The nomotic layer protects them and the brand.

Manufacturing

A manufacturing company uses AI agents to manage supply chain logistics, predictive maintenance, and quality control. The nomotic layer ensures operational safety and compliance.

• Pre-action authorization evaluates maintenance decisions before they are executed. An agent recommending equipment shutdown for preventive maintenance must demonstrate that the recommendation aligns with safety protocols and production schedules.
• Cross-system governance monitors interactions between agents. A procurement agent ordering parts, a logistics agent scheduling delivery, and a maintenance agent planning installation must operate within coordinated authority. The nomotic layer ensures the combined workflow doesn’t create conflicts or compliance gaps.
• Explicit boundaries define what agents can automate versus what requires human approval. Routine reorders proceed automatically. Orders exceeding cost thresholds or involving new suppliers require human authorization.
• Accountable governance maintains clear responsibility chains. When a quality issue emerges, the nomotic layer provides traceability from the defect back through every agent decision and human approval that contributed to it.

The agentic layer optimizes operations. The nomotic layer ensures those operations remain safe and accountable.

Insurance

An insurance company deploys AI agents for claims processing, underwriting assistance, and fraud detection. The nomotic layer ensures fair treatment and regulatory compliance.

• Ethical evaluation assesses claims decisions for fairness. Before an agent denies a claim, the nomotic layer verifies that the decision can be justified and doesn’t reflect prohibited bias.
• Transparent reasoning ensures decisions are explainable to regulators and customers. The nomotic layer requires that every denial include a clear rationale traceable to policy terms.
• Dynamic authorization responds to claim complexity. Straightforward claims within established patterns proceed with minimal friction. Claims with unusual characteristics trigger additional review.
• Human accountability ensures that automated decisions trace to human-approved policies and that exceptions route to human adjusters with full context.

The agentic layer processes claims efficiently. The nomotic layer ensures those decisions are fair and defensible.

Getting Involved

How do I get involved with Nomotic AI?

Just start using it.

There’s no application process, no certification required, and no permission to obtain. If the terminology helps you describe what you’re building, use it. If the principles guide your architecture decisions, apply them.

Practical ways to participate:

• Use the terminology. Put it in your decks, documentation, and discussions. Reference Nomotic principles when explaining your governance approach. The vocabulary becomes more useful as more people adopt it.
• Develop frameworks. Build your own nomotic frameworks tailored to your organization, industry, or use case. Document what works and what doesn’t.
• Share your approach. I’m building a resource of implementation approaches and would welcome contributions. If you’ve developed a nomotic framework, created nomotic tools, or applied Nomotic principles in a novel way, share it. The concept strengthens through collective experience.
• Challenge and refine. The principles and characteristics are not a fixed doctrine. If something doesn’t work, say so. If something is missing, propose it. The framework improves through use and critique.

Nomotic AI belongs to everyone working on AI governance. What you build with it moves the entire field forward.

Summary

Nomotic AI is a category describing intelligent governance for AI systems. It is a counterpart to Agentic AI’s focus on capability and action.

The terminology provides vocabulary for discussing governance with the same sophistication applied to capability. Nomotic principles, nomotic layers, nomotic frameworks, and nomotic tools all describe different aspects of implementing intelligent, dynamic, runtime, contextual, transparent, ethical, and accountable governance.

The category is open. The terminology is available. The principles provide guidance.

What you build with them is up to you.

---

If you find this content valuable, please share it with your network.

Follow me for daily insights.

Schedule a free call to start your AI Transformation.

Book me to speak at your next event.

Chris Hood is an AI strategist and author of the #1 Amazon Best Seller Infailible and Customer Transformation, and has been recognized as one of the Top 40 Global Gurus for Customer Experience. His latest book, Unmapping Customer Journeys, will be published in 2026.