Table of Contents

Reach SOC 2 Compliance in 6 Weeks or Less.

  /

  / CMMC Enclave: What It Is & How It Works

CMMC Enclave: What It Is & How It Works

Defense contractors handling Controlled Unclassified Information now face a choice that shapes their entire compliance budget: lock down the whole organization, or draw a tight boundary around CUI and protect only that.

The second path is kown as the CMMC enclave. For many companies in the Defense Industrial Base, it is the faster, more affordable, and more operationally sensible route to certification, but only if it is scoped and implemented correctly.

This article explains what a CMMC enclave is, how it differs from enterprise-wide compliance, and what it takes to build one that will actually hold up under assessment.

CMMC Enclave: What It Is, How It Works, and Whether It's Right for You

What Is a CMMC Enclave?

A CMMC enclave is a logically or physically isolated segment of your IT environment where all CUI is processed, stored, and transmitted. Everything inside the enclave boundary is in scope for a CMMC assessment. Everything outside is not.

Think of your company as a building. The enclave is a locked, monitored room inside it. Only specific people are authorized to enter, all activity within the room is logged, and the security controls governing the room are documented and continuously enforced. The rest of the building operates normally, unaffected by the rigorous controls applied inside.

The concept is explicitly supported by DoD guidance. The CMMC Level 2 Scoping Guide states that organizations “may limit the scope of the security requirements by isolating the designated system components in a separate CUI security domain.” That isolation can be achieved through physical separation, logical separation, or a combination of both.

How a CMMC Enclave Differs from Enterprise-Wide Compliance

Enterprise-wide compliance means applying all 110 NIST SP 800-171 controls across your entire organization: every endpoint, every user account, every application that touches any part of your network. That is the default interpretation many contractors start with, and it is expensive. A larger scope means more assets to harden, more users to train, more systems to document, and a bigger, more complex assessment.

An enclave approach inverts the logic. Instead of bringing the whole organization up to CMMC Level 2 standards, you identify the minimum set of systems and users that genuinely need to touch CUI — and you apply full controls to only that subset. The result is a smaller, focused compliance footprint.

The financial difference is real. Published case studies show that well-scoped enclaves reduce CMMC implementation costs by 20 to 45 percent compared to enterprise-wide approaches.

A 40-person manufacturer, for example, reduced its projected CMMC implementation cost from $140,000 to $78,000 by migrating CUI into a cloud-based enclave. The savings compound: fewer assets to secure, fewer people to train, a smaller assessment scope, and lower ongoing maintenance costs year after year.

Physical Separation vs. Logical Separation in a CMMC Enclave

The DoD’s own scoping guidance is clear that security domains may use physical separation, logical separation, or a combination of both. Understanding the difference matters because your choice affects architecture, cost, and how an assessor will evaluate your boundary.

Physical separation means CUI assets live on dedicated hardware, in a separate room or cage, disconnected from general-purpose networks at the cable level. It is the most defensible form of separation, but it also carries higher hardware costs and operational overhead. For some regulated environments — particularly those subject to Level 3 requirements or handling the most sensitive categories of CUI — physical separation may be necessary.

Logical separation uses network segmentation, firewall rules, VLANs, and access controls to isolate CUI assets within a shared physical infrastructure. It is cheaper, faster to implement, and the more common approach for CMMC Level 2 enclaves — but it requires architectural rigor. A VLAN boundary that is not technically enforced, or a firewall rule that permits general IT traffic to reach CUI systems, will not hold up during assessment.

A critical point the DoD has reinforced in its updated FAQ guidance: logical separation must be provable and documented. Saying you have logical separation is not enough. You need enforceable architecture, tested configurations, and the documentation to demonstrate both.

Important: A common mistake is treating logical separation as a policy statement rather than an architectural fact. Assessors will test your boundary controls, not just read your System Security Plan. If traffic can flow between your corporate network and your CUI enclave — even indirectly — the enterprise network may be pulled into scope.

Why CMMC Scoping Matters Before Choosing an Enclave Approach

Scoping is the decision that determines everything downstream: which systems you secure, which employees you train, how much the assessment costs, and how confident you can be that you will pass. Getting it wrong in either direction creates problems.

Over-scoping wastes money. If your compliance boundary includes systems that never touch CUI, you are paying to harden infrastructure that does not need it.

Under-scoping is worse: if CUI flows through systems outside your declared enclave — shared email servers, unmanaged endpoints, a consumer file-sharing tool someone uses informally — your boundary is invalid and your assessment will fail.

NIST SP 800-171 offers a useful framing: organizations “will not want to spend money on cybersecurity beyond what it requires for protecting its missions, operations, and assets.”

Scoping is how you align security investment with actual risk. Every asset you can legitimately keep out of scope is a saving.

How to Scope a CMMC Enclave

Scoping starts with a single question: where does CUI actually go in your environment?

The answer is usually more distributed than people expect. CUI flows through email. It lands in shared drives, project management tools, collaboration platforms, and sometimes personal devices. Before you can define an enclave, you need to map all of it.

The DoD scoping process works through asset categories: CUI Assets (systems that directly process, store, or transmit CUI), Security Protection Assets (systems that enforce security functions for CUI assets), Contractor Risk Managed Assets, Specialized Assets (IoT, OT, test equipment), and Out-of-Scope Assets. Only Out-of-Scope Assets can be excluded from assessment — and to qualify, they must be provably isolated from CUI flows.

The key discipline is minimization. The question is not just “which assets handle CUI?” but “which assets must handle CUI?”

Every workflow you can redesign to keep CUI out of a system is a legitimate scope reduction. Route CUI through a dedicated platform. Use a controlled collaboration tool that lives inside the enclave. Stop emailing CUI through your corporate mail server if you can route it through an enclave-resident system instead.

Pro Tip: Conduct a CUI data flow analysis before drawing any boundary

Interview the teams that work on DoD contracts, pull network logs, and review file-sharing configurations. CUI often travels through channels that IT is unaware of — personal email copies, consumer cloud sync, third-party tools with broad file access. Find those flows before your assessor does.

What Does It Mean to Isolate a CMMC Enclave?

Isolation is not a single control. It is a set of architectural decisions that collectively prevent CUI from leaking outside the boundary and prevent unauthorized access from entering.

A properly isolated enclave enforces strict network segmentation: CUI systems sit on a separate network segment with firewall rules that permit only authorized traffic. Identity and access management is enclave-specific: users authenticate through multi-factor authentication to enter the enclave, and access rights are role-based and documented.

Data in transit is encrypted. Data at rest is encrypted. Every access event is logged to an enclave-resident SIEM, and alerts are configured to detect anomalous behavior.

The enclave also needs a System Security Plan that specifically describes its boundaries, the controls in place, how CUI flows within it, and how it interfaces with any external systems. The SSP is not optional and not generic — a copy-pasted enterprise SSP that does not accurately reflect the enclave architecture will be flagged immediately by a C3PAO.

External service providers that touch CUI from within the enclave are also in scope. Your cloud storage provider, managed security vendor, or identity platform all have to meet applicable requirements — typically FedRAMP Moderate authorization at a minimum.

 

Strategic Benefits of Using a CMMC Enclave

Reduce Compliance Scope and Complexity

The most immediate benefit is scope control. When your assessment boundary covers 10 workstations instead of 100, the entire compliance effort shrinks proportionally.

There are fewer endpoints to harden, fewer configurations to document, fewer training requirements to track, and a smaller surface area for an assessor to examine.

For organizations where DoD contracts represent a portion of total revenue rather than the whole business, the enclave approach makes it possible to achieve full CMMC compliance without forcing the commercial side of the company through unnecessary security overhead.

Strengthen CUI Data Protection

Paradoxically, a well-scoped enclave often produces stronger actual security than a sprawling enterprise-wide implementation. When controls are concentrated on a small, purpose-built environment, they can be implemented with precision and maintained rigorously.

Every access event is logged. Every configuration is documented. Patch management, vulnerability scanning and penetration testing are enclave-specific and can be governed tightly. The alternative — applying 110 controls loosely across a large organization — frequently produces compliance theater rather than genuine security.

Insider Note: Defense contractors who pursue enterprise-wide compliance without scoping discipline often end up with a large assessment boundary that is technically non-compliant at the edges. Enclaves, when properly designed, tend to produce cleaner assessments precisely because the scope is manageable enough to implement correctly.

Save on Compliance Costs

CMMC implementation costs scale directly with scope. A well-designed enclave might cover 20 workstations instead of 200, require training 15 people instead of your entire workforce, and reduce the assessment to a focused, manageable exercise.

The ongoing operational savings compound: monitoring, patch management, access reviews, incident response, and annual documentation updates all get cheaper when the scope is smaller. For small and mid-sized defense contractors, the enclave approach is often the difference between compliance being financially viable and not.

Reach SOC 2 Compliance in 6 Weeks or Less

Schedule Your Free SOC 2 Assessment Today

CMMC Enclave vs. Enterprise-Wide Compliance: Key Factors to Consider

The right approach depends on your organization’s specific situation. Neither is universally better. The table below maps the key decision factors against each model.

Factor

Enclave Approach

Enterprise-Wide

DoD contracts as % of revenue

Low to moderate

High

Proportion of staff handling CUI

Minority

Majority or all

Existing IT infrastructure

Mixed commercial/personal

Already segmented or managed

Compliance budget

Constrained

Flexible

Timeline to certification

Shorter

Longer

Operational disruption

Minimal for most staff

Affects entire organization

Long-term scalability

May require redesign if CUI scope grows

Scales more naturally

One nuance that organizations frequently miss: if you map your CUI flows carefully and discover that most of your staff actually does touch CUI — across shared email, shared drives, shared project tools — you may find the enclave approach does not save as much as expected.

When the boundary ends up including most of the organization anyway, the operational complexity of maintaining two separate environments may outweigh the cost savings.

On the other hand, when DoD contracts represent a defined, bounded portion of your business, forcing your entire organization into compliance-grade infrastructure is genuinely wasteful. As one experienced practitioner put it, requiring everyone to wear body armor because some employees work in a secure area makes no operational sense.

 

Pros and Cons of an Enclave-Based Security Approach

The enclave model offers a focused, cost-efficient path to certification. Its advantages are real: smaller scope, lower implementation and assessment costs, less operational disruption to non-CUI staff, faster time to certification, and tighter actual security controls on CUI. For small and mid-sized contractors, these benefits are often decisive.

Its limitations are equally real. Enclaves introduce the complexity of managing two parallel environments. Staff who work across DoD and commercial lines must operate in both. CUI boundaries can drift over time as workflows evolve, requiring active governance to keep the scope accurate. And if the boundary is not correctly defined and technically enforced from the start, the enclave can create a false sense of security while leaving CUI exposed in systems that were never properly accounted for.

 

Pros and Cons of Enterprise-Wide Compliance

Enterprise-wide compliance eliminates the boundary management problem. Everyone operates under the same security posture. There is no risk of CUI leaking outside a declared enclave because there is no separate enclave to maintain. For organizations where CUI handling is genuinely pervasive, this approach may actually be simpler in the long run.

The downside is cost and organizational impact. Applying 110 NIST SP 800-171 controls to an entire organization — including staff who never touch DoD data — drives up technology spend, training requirements, and assessment complexity. It also tends to produce friction with business units that find compliance controls disruptive to commercial workflows.

 

Real-World Scenarios: Where Do You Fit?

A 30-person aerospace manufacturer where five engineers handle CUI on dedicated systems is an obvious enclave candidate. The compliance work is contained, the boundary is defensible, and the rest of the company is unaffected.

A 150-person defense services firm where almost every employee touches CUI as part of contract delivery is a poor enclave candidate. Attempting to enclave in that environment typically produces a boundary that either includes most of the organization anyway, or excludes systems that genuinely need to be in scope — both outcomes are costly in different ways.

A company with a mixed business — commercial software on one side, DoD contracts on the other — is the classic enclave use case. The CUI work can be isolated into a dedicated environment, the commercial side operates normally, and compliance costs are proportionate to the DoD work.

 

CMMC Enclave Models: Hybrid vs. Cloud-Only

Hybrid Boundary Model

The hybrid model combines on-premises infrastructure with cloud services. CUI assets may live partly in a dedicated on-premises segment — a separate VLAN, a physically distinct server, a hardened workstation cluster — with cloud services layered in for specific functions like email, file sharing, or SIEM. This model suits organizations with existing on-premises infrastructure that they cannot or do not want to fully migrate.

The complexity in a hybrid model lies at the boundary between on-premises and cloud components. Every integration point is a potential scope expansion. Identity federation, email routing, file sync configurations, and network connectivity between on-prem and cloud environments all need to be evaluated and documented. Boundary creep is the primary risk.

Cloud-Only Enclave Model

The cloud-only model places the entire enclave in a FedRAMP-authorized cloud environment — typically Microsoft Azure Government GCC High, or AWS GovCloud. Users access CUI through virtual desktop infrastructure or web-based applications. No additional on-premises hardware is required. Isolation is logical and identity-based rather than physical.

This model is increasingly common and well-supported. Microsoft GCC High, in particular, covers the majority of the technical controls required by NIST SP 800-171, which reduces the implementation burden significantly.

The cloud-only approach also benefits from the cloud provider’s inherited controls, which reduces the number of controls an organization must implement and document independently.

Comparing Cost and Complexity Between Models

Cloud-only enclaves typically have lower upfront infrastructure costs, faster deployment timelines (16 to 20 weeks is a common range for a managed cloud enclave), and lower ongoing hardware maintenance overhead.

Hybrid models may suit organizations with legacy on-premises systems that cannot be easily migrated, or those with specific data handling requirements that preclude cloud hosting.

The total cost of compliance over three to five years often favors cloud-only enclaves for small and mid-sized contractors, because the operational burden of maintaining physical infrastructure and the associated documentation is eliminated.

Worth Knowing: Using a FedRAMP-authorized cloud environment

When using a FedRAMP-authorized cloud environment as your enclave platform, the cloud service provider's FedRAMP authorization package documents which controls are inherited versus shared-responsibility. Understanding that boundary is essential to scoping your own SSP correctly. Do not assume all 110 controls are covered by the cloud provider — many are shared or customer-managed.

How to Create a CMMC Enclave Step-by-Step

How to Create a CMMC Enclave: Step-by-Step

Step 1: Discovery and Planning

Begin by mapping every location where CUI currently exists: which systems store it, which users access it, how it moves between systems and people, and where it exits the organization to flow to subcontractors or government systems.

This is not a one-hour exercise. It requires interviews with contract-facing teams, IT system reviews, log analysis, and a careful look at informal channels like personal email forwarding and consumer file-sharing tools.

From that map, identify the minimum viable set of assets that genuinely need to handle CUI. That minimum defines your target enclave. Document your findings and the decisions that follow — the justification for each scoping decision will become part of your assessment evidence.

A structured pre-assessment gap review at this stage can surface control weaknesses before you commit to an architecture.

Step 2: Design Your Enclave Boundary

With your CUI map in hand, design the technical architecture that will enforce the boundary. Define which users are in scope. Identify which systems need to be included. Select the cloud or on-premises platform that will host your CUI environment. Plan your network segmentation, access controls, authentication requirements, logging, and monitoring architecture.

Also plan for data flow into and out of the enclave. CUI that enters through email needs a compliant email system inside the boundary. File transfers to and from government clients need encrypted, auditable channels. Subcontractors who receive CUI from you become external service providers in scope — assess their status and document it.

Step 3: Deployment

Implement the enclave architecture. Deploy the platform, configure network segmentation, enforce multi-factor authentication for all enclave access, apply encryption at rest and in transit, implement a SIEM for logging and alerting, and begin vulnerability scanning and penetration testing. Apply NIST SP 800-171 controls systematically, documenting each implementation decision in your System Security Plan.

Migrate existing CUI assets into the enclave and retire or isolate the out-of-scope systems they previously lived on. Train enclave users on their specific responsibilities under CMMC compliance.

Step 4: Validation

Before engaging a C3PAO, conduct a formal gap assessment against all 110 NIST SP 800-171 controls. Document any gaps in a Plan of Action and Milestones (POA&M) and remediate what you can. Your SPRS score — the self-assessed score submitted to the Supplier Performance Risk System — must reflect your actual implementation status, not an aspirational one.

An internal audit at this stage, or engagement with a Registered Practitioner Organization (RPO), can identify boundary definition issues and control gaps before a formal assessment, when corrections are still inexpensive.

Step 5: Ongoing Operations and Monitoring

Certification is not a one-time event. Maintaining CMMC Level 2 status requires continuous monitoring and operational discipline: daily monitoring tasks, weekly log reviews, monthly access audits, quarterly vulnerability scans, and annual control reviews. Personnel changes, new software deployments, and contract awards can all trigger additional compliance work.

Establish governance processes to keep your enclave boundary accurate over time. As your business evolves — new DoD contracts, new collaboration tools, staff changes — your CUI data flow can shift, pulling new systems or people into scope without anyone noticing. An annual scoping review is the minimum; quarterly is better. Tools like compliance automation tools like Drata and Vanta can help surface drift in control coverage and reduce the manual burden of evidence collection over time.

Reach SOC 2 Compliance in 6 Weeks or Less

Schedule Your Free SOC 2 Assessment Today

Is a CMMC Enclave Right for Your Business?

Business Use Cases for a CMMC Enclave

The enclave approach fits well in specific situations. Manufacturing companies with a dedicated engineering team handling contract specifications and technical drawings — while the rest of the business handles commercial operations — are strong candidates.

Professional services firms where a defined project team handles DoD engagements while other staff work on commercial clients benefit from the same dynamic. Small and mid-sized contractors with limited IT budgets who cannot afford to bring their entire organization up to CMMC certification standards often find the enclave the only financially viable path.

Who Should Choose an Enclave Approach vs. Enterprise-Wide Compliance

Choose an enclave if: your DoD contracts represent a distinct, bounded portion of your business; the number of staff who genuinely need CUI access is a minority of your workforce; your existing commercial IT environment would require substantial overhaul to meet CMMC requirements; and you want to achieve certification on the fastest, most cost-controlled timeline.

Choose enterprise-wide compliance if: most of your revenue comes from DoD contracts; CUI handling is distributed broadly across your organization; your existing IT environment is already heavily managed and segmented; or you anticipate significant growth in DoD contract volume that would expand the enclave boundary to near-enterprise scale anyway.

Pro Tip: If you are genuinely uncertain which approach applies to you, start with the data flow analysis before making any architecture decisions. Where CUI actually lives and moves in your organization is the only reliable basis for the choice. Architectural decisions made before that analysis is complete tend to be either over-engineered or under-scoped.

 

Future-Proofing Your Enclave for CMMC 2.0 and Beyond

The CMMC program is now in active enforcement. The 32 CFR Part 170 rule became effective in December 2024, and the 48 CFR acquisition rule requiring CMMC in DoD contracts became effective in November 2025. The program is no longer a future obligation — it is a present contractual requirement for defense contractors.

Enclaves built to CMMC Level 2 standards today should be designed with adaptability in mind. NIST SP 800-171 Rev. 3 introduced updates to the control framework, and the DoD has been progressively tightening its interpretive guidance on scoping through quarterly FAQ updates.

An enclave that was defensible under prior guidance may require adjustment as DoD clarifies expectations around logical separation, external service provider scope, and boundary documentation.

Build your enclave on a FedRAMP-authorized platform where possible. Document every architectural decision and the reasoning behind it. Engage a C3PAO or RPO to review your scoping logic before it is tested in the assessment. And treat your System Security Plan as a living document, not a compliance artifact to be filed and forgotten.

The organizations that will maintain CMMC certification most sustainably are not necessarily those with the most sophisticated technology. They are the ones with accurate scope definition, rigorous documentation, and the operational discipline to keep their enclave current as their business evolves. If you want to accelerate that process, the Compliance Accelerator Program (learn more) is designed to give defense contractors a structured, supported path from discovery to assessment readiness.

What is a CMMC enclave, and why would I need one?

A CMMC enclave is an isolated segment of your IT environment — logically or physically separated from the rest of your network — where all CUI is processed, stored, and transmitted.

You need one if you want to limit your CMMC assessment scope to only the systems and users that genuinely handle CUI, rather than subjecting your entire organization to the full set of NIST SP 800-171 controls.

It reduces compliance cost, assessment complexity, and operational disruption for the parts of your business that do not touch DoD data.

No. CMMC compliance does not require you to use an enclave. It requires you to implement the controls appropriate to your certification level across all systems within your CMMC assessment scope.

The enclave is a scoping strategy, not a program requirement. You can pursue CMMC compliance with or without one — but for most small and mid-sized contractors, an enclave is the most practical way to manage scope and cost.

Yes, and cloud-hosted enclaves are increasingly the preferred model. FedRAMP-authorized cloud environments — including Microsoft Azure Government GCC High and AWS GovCloud — provide a compliant platform that inherits a significant portion of the NIST SP 800-171 controls.

Users access the enclave through VDI or web applications, no dedicated on-premises hardware is required, and the cloud provider’s authorization documentation supports your SSP. Cloud-only enclaves typically deploy faster and carry lower ongoing operational overhead than hybrid models.

Timeline varies based on your starting point, the complexity of your CUI environment, and whether you are building on an existing compliant platform or starting from scratch.

A cloud-based managed enclave built on a FedRAMP-authorized platform by an experienced provider typically deploys in 16 to 20 weeks from discovery to assessment readiness. Hybrid models with significant on-premises components may take longer.

Organizations with poorly documented CUI flows or fragmented existing infrastructure should budget additional time for the discovery and design phases.

Cost depends heavily on scope — the number of users, systems, and the platform chosen. A small enclave serving 10 to 20 users built on a cloud platform with a managed service provider can range from roughly $30,000 to $80,000 for initial implementation, with ongoing managed service costs thereafter.

Larger or more complex enclaves, or those requiring significant on-premises build-out, will cost more.

The relevant comparison is not the absolute cost but the cost relative to enterprise-wide compliance: well-scoped enclaves consistently deliver 20 to 45 percent cost reductions compared to protecting an organization’s entire IT estate.

The enclave model reduces the blast radius of a breach by concentrating CUI into a defined, heavily monitored environment. If a breach occurs within the enclave, the CUI at risk is limited to what is inside the boundary — rather than being scattered across a broad enterprise environment. That said, a breach of CUI is a significant event regardless of enclave architecture.

DFARS 252.204-7012 requires notification to the DoD within 72 hours of discovering a cyber incident, and CUI compromised in the breach must be reported. The enclave’s logging and continuous monitoring capabilities are what make rapid detection and response possible.

For most contractors, yes. The scoping analysis, architecture design, SSP documentation, and pre-assessment gap review are all areas where experienced guidance reduces both cost and risk. Registered Practitioner Organizations (RPOs) are authorized to provide CMMC consulting and can help you define your enclave correctly before you invest in implementation.

Working with an RPO also helps you avoid the most common and costly mistake in enclave design: drawing a boundary that looks defensible on paper but fails under technical scrutiny during assessment.

If you would like to explore how Axipro approaches CMMC enclave scoping and implementation, contact our team for an initial assessment.

Axipro Author

Picture of Pedro Dias

Pedro Dias

Pedro has been writing online for over 10 years. With experience in all things programming, cyber security, and compliance, he is our editor-in-chief at Axipro.

Blog Highlights

Explore More Articles

For the past two years, enterprise AI risk conversations have centered on a familiar set of concerns: model bias, hallucination, data privacy, and dependency on third-party models. These are real risks, and most organizations now run some version of a governance program to manage them. But something has shifted. Organizations are no longer just deploying AI that generates content for a human to review. They’re deploying AI that acts. Agents now plan multi-step tasks, call APIs, move data between systems, execute transactions, and coordinate with other agents, often with no human checkpoint in the loop. That shift deserves more than a footnote in the existing AI risk category. It deserves its own line in the risk register: Agentic Autonomy Risk. What Is Agentic AI Risk Management? Agentic AI risk management is the practice of identifying, assessing, and controlling the risks created when AI systems take autonomous action on an organization’s behalf. Where traditional AI governance evaluates outputs (accuracy, bias, privacy), agentic AI risk management governs what agents actually do: the tools they call, the permissions they inherit, and the downstream consequences of their actions. That distinction is the reason existing risk registers struggle with agents, and it’s worth unpacking properly. What Agentic AI Actually Changes Traditional AI systems, even generative ones, are advisory. They produce an output such as a summary, a prediction, a draft email, or a classification, and a human remains the last checkpoint before anything happens in the real world. Agentic AI removes that checkpoint. An agentic system doesn’t just produce an answer. It pursues a goal. It decides which tools to call and in what order, then executes those actions directly against live systems: submitting a purchase order, modifying a database record, sending an external communication, or orchestrating a set of sub-agents to complete a broader workflow. Agentic autonomy is the degree to which a system can plan and execute actions without a human explicitly authorizing each step. It’s a spectrum rather than a binary. At one end, the AI drafts and a human approves every action. At the other, the AI operates within broad guardrails and only escalates exceptions. The further an organization moves along that spectrum, the less its exposure looks like software risk and the more it looks like delegated authority risk, the kind normally reserved for employees, contractors, and automated financial systems. Why Existing Risk Registers Miss Agentic AI Risks Most enterprise risk registers were built on a reasonably safe assumption: a human initiates consequential actions, and the technology around that human behaves deterministically. Agentic AI breaks both halves of that assumption at once. A few specific gaps show up quickly when organizations try to map agentic deployments onto existing categories. Operational risk registers assume process failures come from human error or system outages, not from a system independently choosing an unanticipated path to a stated goal. Cybersecurity risk registers are built around unauthorized external access, while an agent problem usually involves an authorized system taking unauthorized internal actions with its own legitimate credentials. Model risk frameworks, borrowed largely from financial services, evaluate output accuracy rather than action consequences, which matters most when those actions can’t be reversed. And third-party risk assessments treat vendors as static entities, not as autonomous agents that might invoke other vendors’ agents on your behalf. See our guide to the NIST AI Risk Management Framework for how output-focused frameworks are structured. The result is a governance blind spot. An organization can be compliant against its AI policy, its cybersecurity policy, and its vendor risk policy, and still have nobody accountable for the specific risk of a system initiating a harmful sequence of actions before anyone notices. Defining Agentic Autonomy Risk Agentic Autonomy Risk is the risk that an AI system, operating with delegated decision-making and execution authority, takes actions that are harmful, non-compliant, or misaligned with organizational intent before adequate human oversight can intervene. Those actions might happen independently or in coordination with other agents. It deserves standing as a named category alongside cybersecurity, operational, legal, financial, and third-party risk because the loss event itself is different. The harm is a completed action in a live system, and it may be difficult or impossible to reverse. The accountability structure is different too: when an orchestrating agent delegates to sub-agents, responsibility for the outcome gets distributed in ways existing ownership models don’t cleanly capture. So is the detection window. Traditional controls assume a human is positioned to catch an error before it compounds, but an agent can execute dozens of dependent actions faster than any human review cycle. 7 Agentic AI Risk Scenarios to Put on Your Register 1. Unauthorized autonomous decision-making. An agent takes an action within its technical permissions but outside its intended business mandate. It adjusts pricing, approves a refund, or modifies a customer record, and no policy ever explicitly authorized that scenario. 2. Goal misalignment. The agent optimizes for a literal interpretation of its objective in a way that diverges from actual business intent, particularly under ambiguous or adversarial inputs. 3. Multi-agent interactions and cascading failures. One agent’s flawed output becomes another agent’s trusted input. A single error can propagate across a chain of agents faster than anyone can detect it, amplifying the original mistake instead of containing it. 4. Excessive tool or system permissions. Agents get provisioned with broad, standing access “to be safe” rather than scoped, least-privilege access tied to specific tasks. A productivity tool quietly becomes a privilege-escalation path. 5. Regulatory non-compliance. Autonomous actions trigger obligations under data protection, financial services, employment, or sector-specific regulation, and they execute without the compliance review a human-initiated process would normally receive. 6. Explainability and accountability gaps. An autonomous action causes harm and the organization can’t clearly reconstruct why the agent chose that path, or establish whether the business owner, the AI governance function, or the vendor is accountable for the outcome. 7. Autonomous third-party actions. A vendor’s agent, integrated into your environment, takes action on your behalf, or your agent acts against a

A SOC 2 penetration test costs between $1,000 and $30,000 for most companies. A typical SaaS scope, meaning one web application, its API layer, and the cloud infrastructure behind it, usually lands between $2,000 and $20,000. Early-stage startups with a narrow scope can get an auditor-accepted test for $1,000 to $8,000, while enterprises with multiple products and hybrid infrastructure regularly spend $20,000 to $50,000 or more. The spread is wide because “penetration test” covers everything from an automated scan with a cover page to weeks of manual testing by senior engineers. Auditors know the difference, and so do the enterprise customers who asked for your SOC 2 report in the first place. This guide breaks down what drives the price, where the hidden costs sit, and how to buy a test that holds up in fieldwork without overpaying for it. What Is SOC 2 Penetration Testing?​ A SOC 2 penetration test is a simulated attack on your systems, performed by a qualified security professional, scoped to the environment covered by your SOC 2 report. The tester tries to exploit real weaknesses the way an attacker would: broken access controls, injection flaws, misconfigured cloud services, exposed credentials. The output is a report your auditor reads as evidence that your security controls work in practice, not only on paper. That last part matters. A pentest bought for SOC 2 has a second audience beyond your security team. If the report doesn’t map findings to your audit scope, document its methodology, and show remediation, it fails the job you bought it for. We cover the full deliverable in our guide to what a SOC 2-ready VAPT report includes. How Penetration Testing Fits Into SOC 2 Compliance​ SOC 2 is built on the AICPA’s Trust Services Criteria, and the Security category (the Common Criteria) applies to every report. Penetration testing is the standard way to satisfy CC7.1, which expects you to detect and monitor for new vulnerabilities, and it supports CC4.1, which covers ongoing evaluations of whether controls actually function. The AICPA’s points of focus explicitly mention vulnerability scanning and penetration testing as examples of how companies meet these criteria. In practice, the test slots into your audit timeline as an evidence item. Your auditor will ask for the report, check the test date against the audit period, and review how you handled the findings. Remediation is often scrutinized harder than the test itself, because it shows whether your vulnerability management process runs or merely exists. Is Penetration Testing Required for SOC 2?​ Strictly speaking, no. The Trust Services Criteria never use the word “mandatory” about penetration testing. You could theoretically satisfy CC7.1 with vulnerability scanning and strong monitoring alone. In reality, almost every auditor expects one, and skipping it invites two problems. First, your auditor may push back during fieldwork or add exceptions to the report. Second, the enterprise buyers reviewing your SOC 2 report increasingly look for pentest evidence specifically, and a report without it raises questions during procurement. Treat the test as effectively required and budget for it from the start of your SOC 2 compliance checklist. How Much Does SOC 2 Penetration Testing Cost? Typical Price Range for SOC 2 Pen Testing Most companies pay $1,000 to $30,000, with the median engagement for a SaaS business sitting around $12,000 to $15,000. Compliance-focused tests at the lower end of the market start around $1,000 to $5,000. Deep manual testing from established firms runs $10,000 to $30,000. Anything quoted below roughly $3,000 is almost certainly automated scanning packaged as a pentest, which auditors are getting better at spotting. Cost by Company Size (Startup, SMB, Enterprise) Company size is a proxy, not the driver. A 15-person company with three products and a legacy on-prem component will pay more than a 200-person company with one tightly scoped SaaS platform. Testers price effort, and effort follows scope. Cost by Test Type (Network, Web App, API, Cloud, Internal/External) Most SOC 2 engagements bundle two or three of these. The common package for a cloud-native SaaS company is web app plus API plus cloud configuration, which is why the $1,000 to $20,000 band comes up so often. Companies with office networks and internal systems in their audit scope add internal network testing, and the price climbs accordingly. Factors That Influence SOC 2 Penetration Testing Cost Scope and Number of Assets Tested Scope is the single biggest cost driver. Every additional application, API endpoint group, cloud account, or network segment adds testing hours. A pentest priced without a scoping call is a pentest priced on guesswork, and the guess usually favors the vendor. Complexity of Application or Infrastructure​ A simple CRUD app with two user roles tests quickly. A multi-tenant platform with role hierarchies, workflow engines, file processing, and third-party integrations takes far longer, because each of those features creates attack surface a tester has to work through manually. Authentication tiers matter especially: every distinct role needs testing for privilege escalation and cross-tenant data access. Testing Methodology (Black Box, Grey Box, White Box) Black box testing gives the tester nothing but a URL, grey box adds credentials and documentation, and white box adds source code and architecture diagrams. Grey box is the default for SOC 2 and usually the best value, since the tester spends time exploiting rather than discovering. White box costs more upfront but finds deeper issues. Black box sounds rigorous but often wastes paid hours on reconnaissance an attacker would run for free. Depth of Testing and Manual vs. Automated Approaches Automated scanning finds known vulnerability patterns. Manual testing finds business logic flaws, chained exploits, and authorization gaps that no scanner catches, and it’s the part auditors and security-literate customers actually value. The ratio of manual work to automation is the honest explanation for most price differences between two quotes covering the same scope. Tester Credentials and Firm Reputation Senior testers holding OSCP, GPEN, or CREST credentials bill higher rates, and firms with recognized methodologies charge a premium for the credibility their letterhead carries

Two compromised versions of LiteLLM sat on PyPI for roughly 40 minutes on the morning of March 24, 2026. That window was enough to capture secrets from around 434,000 CI/CD pipeline runs across nearly 2,500 organizations, including AWS, Samsung, Cisco, Salesforce, Siemens, and Deloitte. In August, researchers at CloudSEK and Hudson Rock confirmed they had obtained the raw exfiltrated data: a 153GB archive containing 433,909 files of environment variables, cloud keys, Kubernetes secrets, and API tokens harvested live from running pipelines, as covered by Help Net Security’s reporting on the credential archive. If LiteLLM runs anywhere in your stack, or you touch any AI proxy infrastructure at all, you need answers to three things: whether you were exposed, what to rotate first, and whether the rotation you did back in March actually held. That last one matters more than it sounds, because “we rotated everything” has already burned at least one very large company. How the Breach Happened The attack didn’t start with LiteLLM. On March 19, 2026, a threat group called TeamPCP compromised the build pipeline of Trivy, a vulnerability scanner half the industry runs, and pushed a poisoned release. LiteLLM’s own CI pipeline ran Trivy, so the poisoned scanner had legitimate read access to the project’s runner environment. The attackers used that to steal LiteLLM’s PyPI publishing tokens and ship two malicious releases of their own: versions 1.82.7 and 1.82.8. KICS and the Telnyx Python SDK got hit in the same campaign. The payload design is the part worth studying. The malicious package dropped a .pth startup hook into site-packages, so the code ran the moment any Python interpreter started on the machine, whether or not anything imported LiteLLM. From there it harvested environment variables, read local credential files like .aws/credentials and .kube/config, tried to move laterally across Kubernetes clusters, and installed a systemd backdoor dressed up as a generic telemetry service. InfoQ’s coverage of the PyPI compromise put downloads of the compromised release above 40,000. For scale, LiteLLM normally gets downloaded around 3 million times a day. The exfiltration had a nasty fallback, too. According to CloudSEK, stolen data was encrypted and sent to a typosquatted domain, and when that failed, the malware created a public repository inside the victim’s own GitHub account and uploaded the loot as a release asset. Some companies were publishing their own secrets to the open internet and had no idea. Worth Knowing: The malicious code only existed in the PyPI artifacts. The GitHub source repository stayed clean the whole time, so a developer reviewing the code on GitHub saw nothing wrong. Source review isn’t artifact verification. If you don’t check that what the registry serves matches the upstream source, this class of attack is invisible to you. How to Check If You Were Exposed Three checks, from quickest to most involved. 1. Confirm whether the compromised versions ever ran The malicious versions went live on PyPI at 10:39 UTC on March 24, 2026 and got quarantined about 40 minutes later. The project’s advice: treat any install from that day before 16:00 UTC as suspect. Search your lockfiles, pip caches, SBOMs, and container image histories for 1.82.7 and 1.82.8. And check your internal artifact mirrors. An Artifactory or Nexus proxy that cached the bad release in March can keep serving it internally long after PyPI pulled it. Keep the .pth mechanism in mind when you scope this. The question isn’t “which applications import LiteLLM,” it’s “which machines had the package installed at all,” because every Python process on an infected machine triggered the payload. 2. Hunt for persistence Rotation is pointless if the attacker still has a foothold. Check developer machines, CI runners, and containers for unauthorized .pth files in site-packages and for suspicious systemd units, especially anything posing as a system telemetry service. And review activity from March 24 onward, not just the 40-minute window. Persistence is there so the access outlives the infection. Pro Tip: Don’t limit the persistence hunt to live machines. Base container images rebuilt in late March may have baked the payload into every image derived from them since. Scan your image registry for the affected LiteLLM versions and for unexpected .pth files, then trace which running workloads came from flagged images. 3. Check whether your secrets are in the dump Hudson Rock has published a domain lookup tool and is running ethical disclosures for affected organizations, and CloudSEK maintains a high-confidence victim list. Use them, but know their limits. Attribution in this dataset is genuinely hard. One dump with a siriusxm.com committer email actually traced, through its self-hosted GitLab endpoints, to AdsWizz, a SiriusXM subsidiary. And a large share of the dumps are generic pipeline configurations with no identifying domain, email, or server name at all. Absence from a victim list is not evidence of absence. If your pipelines ran the compromised versions, assume exposure no matter what a lookup tool tells you. What to Rotate, in What Order The guidance from both research teams is blunt: treat every secret the LiteLLM environment could reach as compromised. That covers secrets on disk, in memory, injected into CI jobs, and anything retrievable through instance metadata services. Work down by blast radius: Priority Credential type Why it comes first 1 Cloud IAM keys (AWS, GCP, Azure) Direct control of infrastructure, data stores, and billing. This is where attackers monetize fastest. 2 GitHub and GitLab PATs, package publishing tokens These let an attacker poison your releases and turn your company into the next link in the supply chain. 3 Kubernetes service account tokens and kubeconfigs Lateral movement across clusters was built into the payload, not a theoretical risk. 4 Database passwords and third-party API keys Dumped in plain text in the archive, often with no attribution, so nobody will warn you they leaked. 5 AI provider API keys Billing abuse, quota theft, and access to whatever data flows through your LLM routing layer. One word matters more than the rest of this article: revoke, don’t just rotate. That