SOC 2 CC6.1: What It Requires, How Auditors Test It, and the Evidence to Prepare
SOC 2 CC6.1 explained by a practicing auditor: all 13 points of focus, the controls that meet them, the evidence to prepare, and the exceptions we see most.
On this page · 13 sections
- SOC 2 CC6.1 is the foundational logical access criterion. It asks whether the entity has implemented logical access security software, infrastructure, and architectures that protect information assets from security events.
- The 2022 revised points of focus list 13 considerations under CC6.1, from asset inventory and authentication to network segmentation, encryption, and key protection. Two of them apply only when Confidentiality or Privacy is in scope.
- Points of focus are not a checklist. Neither MFA nor encryption at rest is a named requirement of SOC 2, but in practice an auditor will expect both wherever your risk assessment says the data or access path is sensitive.
- Most CC6.1 exceptions come from gaps in coverage, not missing tools: a production database outside SSO, a shared admin account, or an asset inventory that does not match the system description.
What is SOC 2 CC6.1?
SOC 2 CC6.1 is the first criterion in the CC6 Logical and Physical Access Controls series of the AICPA 2017 Trust Services Criteria. It requires a service organization to implement "logical access security software, infrastructure, and architectures" over its protected information assets so that those assets are protected from security events and the organization can meet its objectives. In plain terms, CC6.1 asks whether the technical foundation of access control exists: identity, authentication, network boundaries, and encryption.
CC6.1 is part of the common criteria, so it applies to every SOC 2 examination regardless of which Trust Services Categories are in scope. The criterion itself has not changed since 2017. What changed in 2022 is the set of points of focus that illustrate how an organization might meet it.
The criterion is deliberately broad. That breadth is why CC6.1 often maps to more controls than any other criterion in a SOC 2 report. In the reports I review, it is common to see eight to fifteen controls mapped to CC6.1, covering the identity provider, password and MFA configuration, network security groups, encryption settings, and key management.
Where does CC6.1 fit within the CC6 series?
CC6.1 sets up the access control architecture, and the criteria that follow it govern how that architecture is used. Understanding the boundaries between them prevents the most common mapping mistake, which is putting every access control under CC6.1.
| Criterion | What it addresses | Typical controls |
|---|---|---|
| CC6.1 | Logical access security software, infrastructure, and architecture | SSO and MFA, network segmentation, encryption, key management, asset inventory |
| CC6.2 | Registering and authorizing users before issuing credentials, and removing credentials | Access request approvals, onboarding and offboarding |
| CC6.3 | Authorizing, modifying, and removing access based on roles, least privilege, and segregation of duties | Role design, role-change access updates, access reviews |
| CC6.4 | Restricting physical access to facilities and protected information assets | Badge access, data center controls, cloud provider SOC 2 review |
| CC6.5 | Discontinuing protections over physical assets only after data is unrecoverable | Media sanitization, certificates of destruction |
| CC6.6 | Protecting against threats from sources outside system boundaries | Firewalls, WAF, VPN, external-facing hardening |
| CC6.7 | Restricting the transmission, movement, and removal of information | TLS in transit, DLP, removable media restrictions |
| CC6.8 | Preventing or detecting unauthorized or malicious software | Endpoint protection, software installation restrictions |
A useful rule of thumb: CC6.1 answers "is the access control system built correctly?" CC6.2 and CC6.3 answer "is it being used correctly for each person?" Encryption of data in transit generally sits under CC6.7, not CC6.1. Encryption at rest and key management sit under CC6.1.
What are the CC6.1 points of focus?
The CC6.1 points of focus are 13 illustrative considerations published in the 2017 Trust Services Criteria (With Revised Points of Focus, 2022). Eleven apply to every SOC 2. Two apply only when the Confidentiality or Privacy category is in scope. The table below summarizes each point of focus in practical terms and shows a control that commonly addresses it.
| # | Point of focus | What it means in practice | Example control |
|---|---|---|---|
| 1 | Identifies and Manages the Inventory of Information Assets | You know which systems, data stores, and devices hold in-scope data, and someone owns that list | Asset inventory maintained in the GRC platform or CMDB and reviewed periodically |
| 2 | Assesses New Architectures | New system architectures are security-assessed before they go live | Architecture or design review with security sign-off for new services |
| 3 | Restricts Logical Access | Access to assets is restricted through access control software and rule sets | Production access only through SSO-enforced roles |
| 4 | Identifies and Authenticates Users | People, infrastructure, and software are identified and authenticated before accessing assets, locally or remotely | Unique user IDs; service accounts tied to named owners |
| 5 | Considers Network Segmentation | Network segmentation separates unrelated portions of the information system | Separate VPCs or accounts for production, staging, and corporate |
| 6 | Manages Points of Access | Points of access by outside parties, and the data flowing through them, are identified and managed | Inventory of external integrations, API gateways, and vendor access paths |
| 7 | Restricts Access to Information Assets | Access to data, software, functions, and assets is restricted through access control software | Database access restricted to approved roles; customer data segregated by tenant |
| 8 | Manages Identification and Authentication | Identification and authentication requirements are defined, documented, and managed | Password and MFA standard enforced through the identity provider |
| 9 | Manages Credentials for Infrastructure and Software | New infrastructure and software are registered and authorized before receiving credentials, and credentials are removed when no longer needed | Secrets manager with ownership, rotation, and revocation of keys |
| 10 | Uses Encryption to Protect Data | Encryption supplements other measures to protect data at rest, where risk assessment deems it appropriate | Encryption at rest enabled on databases, storage buckets, and backups |
| 11 | Protects Encryption Keys | Encryption keys are protected during generation, storage, use, and destruction | Cloud KMS with restricted key administration and rotation |
| 12 | Restricts Access to and Use of Confidential Information for Identified Purposes (Confidentiality) | Logical access to confidential information is limited to identified purposes | Access to confidential customer data limited to support roles with a ticketed reason |
| 13 | Restricts Access to and the Use of Personal Information (Privacy) | Access to personal information is restricted consistent with privacy commitments | Role-based restrictions on PII fields; access logging for personal data |
Two points are worth stressing.
First, the AICPA states that points of focus are not requirements. A service organization is not required to address each one, and the auditor does not issue an opinion on them. They exist to help management design controls and to help the auditor judge whether the criterion is met. If a point of focus does not apply to your system (for example, you have no external points of access beyond your own application), you do not need a control for it.
Second, "Assesses New Architectures" is one of the newer considerations and is frequently missing from first-year control sets. If your engineering team ships new services through a design review, make sure that review includes a security step and leaves a record.
Is MFA required for SOC 2 under CC6.1?
MFA is not a named requirement in SOC 2. The criteria and points of focus do not say "multi-factor authentication must be used." CC6.1 requires that you implement logical access security appropriate to your objectives and risks, and the points of focus ask you to define and manage identification and authentication requirements.
In practice, the answer for almost every SaaS company is still yes. If your risk assessment identifies account takeover of an administrator or engineer as a risk to customer data, a password-only control is very hard to defend as adequate. Enterprise customers reading your report will also look for it. The defensible position is:
- MFA enforced at the identity provider for all workforce users.
- MFA or equivalent strong authentication for all access to production infrastructure and the cloud console.
- Documented exceptions, if any, with compensating controls and an owner.
What I test is not whether MFA "exists" but whether it is enforced. An identity provider policy that requires MFA for one group while a second group of contractors bypasses it is an exception waiting to happen.
Is encryption at rest required for SOC 2?
Encryption at rest is not mandatory on its own terms. The point of focus says the entity uses encryption to supplement other measures used to protect data at rest when such protections are deemed appropriate based on assessed risk. That wording puts the decision in your risk assessment.
For a SaaS company storing customer data in a managed cloud database, enabling encryption at rest is typically a configuration setting with negligible cost. Choosing not to enable it would need a documented risk rationale that an auditor and your customers would accept. Most organizations simply encrypt databases, object storage, and backups, then protect the keys.
Key protection matters as much as encryption. If every engineer can administer the keys that protect production data, the encryption control is weaker than it looks. A strong design separates key administration from key use, restricts who can schedule key deletion, and logs key usage.
How do auditors test CC6.1?
Auditors test CC6.1 through a combination of inquiry, inspection, observation, and sometimes reperformance. Inquiry alone is never sufficient. The approach differs between a Type 1 and a Type 2 examination.
| Examination | What the auditor evaluates | Typical procedures |
|---|---|---|
| SOC 2 Type 1 | Whether controls are suitably designed and implemented as of a specific date | Inspect policies; inspect identity provider, network, and encryption configurations as of the report date; observe a login requiring MFA |
| SOC 2 Type 2 | Whether controls operated effectively throughout the period, often 3 to 12 months | Inspect configurations at multiple points or test the change management controls that protect them; sample periodic reviews; inspect logs and exception records |
Many CC6.1 controls are automated configurations: an MFA policy, a password policy, an encryption setting. For automated controls, auditors commonly test the configuration and then rely on effective IT general controls, especially change management under CC8.1, to conclude that the configuration did not change without authorization during the period. That is why a weak change management process can undermine the testing approach for CC6.1 even when the configuration itself is correct. The exact approach is a matter of the service auditor's professional judgment under AT-C Section 205.
Manual or recurring CC6.1 controls, such as a quarterly asset inventory review or a periodic firewall rule review, are tested by sampling occurrences across the period. See our guide to SOC 2 sample sizes for how auditors typically size those samples.
What evidence do auditors request for CC6.1?
Auditors request evidence that shows the access control architecture exists, is configured as described, and covers every in-scope system. The list below reflects common requests. Your auditor's list will depend on your controls and system.
| Evidence | What it demonstrates |
|---|---|
| Asset inventory export with owners and in-scope flags | Points of focus 1 and 6: you know what you are protecting |
| System description boundary and data flow diagram | Inventory and architecture match what the report describes |
| Identity provider configuration screenshots or exports (MFA, password, session settings) | Points of focus 3, 4, and 8 |
| List of applications connected to SSO, compared with the in-scope system inventory | Coverage: no in-scope system bypasses central authentication |
| Cloud IAM policy and role listings, including root or owner accounts | Restricted logical access and protection of the highest-privilege accounts |
| Network diagram and security group or firewall rule exports | Segmentation and managed points of access |
| Encryption-at-rest settings for databases, storage, and backups | Point of focus 10 |
| KMS key policies and key administrator list | Point of focus 11 |
| Secrets manager inventory with owners and rotation settings | Point of focus 9 |
| Design or architecture review records for new services in the period | Point of focus 2 |
A practical tip: when you provide configuration evidence, include the system name, the date, and the account or tenant identifier in the screenshot or export. Evidence without those details often triggers a follow-up request.
What are the most common CC6.1 exceptions?
The most common CC6.1 exceptions come from incomplete coverage rather than from missing security tools. The patterns below appear repeatedly across SOC 2 Type 2 engagements.
1. An in-scope system outside SSO and MFA
A production database, a legacy admin panel, or a monitoring tool uses local accounts that sit outside the identity provider. The SSO control is designed correctly but does not cover the whole system. Fix this by comparing your SSO application list with your in-scope inventory before the audit, and either connect each system or document a compensating control.
2. Shared or generic administrator accounts
A shared "admin" login for a vendor portal or a cloud root account used day to day breaks the requirement to identify and authenticate individual users. Root and break-glass accounts should be locked away, protected with hardware MFA, monitored, and used only through a documented procedure.
3. Inventory that does not match reality
The asset inventory lists twelve production services, the system description lists fifteen, and the cloud console shows nineteen. Auditors notice. Reconcile all three before fieldwork.
4. Service accounts and API keys with no owner
Long-lived API keys created by engineers who have since left, or service accounts with administrator rights and no named owner, fail the credential management point of focus. A secrets inventory with owners and a rotation or revocation process addresses this.
5. Encryption claimed but not enabled everywhere
The policy says all data at rest is encrypted, but one storage bucket, a snapshot, or a staging copy of production data is not. Automated configuration monitoring catches this early.
6. Overly permissive network rules
Security groups allowing inbound access from anywhere to administrative ports are a frequent finding. They straddle CC6.1 and CC6.6, and auditors will ask why they exist and who approved them.
How does CC6.1 apply to cloud-native companies?
CC6.1 applies to cloud-native companies through the shared responsibility model. Your cloud provider operates the physical and virtualization layers. You are responsible for identity, configuration, network rules, and encryption settings within your cloud accounts.
In a SOC 2 report, the cloud provider is usually a subservice organization presented under the carve-out method. The provider's controls are excluded from your report, and the auditor expects you to monitor them, typically by obtaining and reviewing the provider's own SOC 2 report each year. Your CC6.1 controls cover the layers you configure. If your report relies on controls you expect your customers to perform, such as managing their own users in your application, those belong in the report as complementary user entity controls.
How does CC6.1 map to ISO 27001:2022?
CC6.1 maps to several ISO/IEC 27001:2022 Annex A controls. The table below shows close equivalents. It is a practitioner mapping for orientation, not an official crosswalk, and a single control can partially satisfy more than one requirement.
| CC6.1 theme | Closest ISO 27001:2022 Annex A controls |
|---|---|
| Asset inventory | A.5.9 Inventory of information and other associated assets |
| Access control policy and restriction | A.5.15 Access control; A.8.3 Information access restriction |
| Authentication | A.5.17 Authentication information; A.8.5 Secure authentication |
| Privileged access | A.8.2 Privileged access rights |
| Network segmentation | A.8.20 Networks security; A.8.22 Segregation of networks |
| Encryption and key management | A.8.24 Use of cryptography |
| New architectures | A.8.27 Secure system architecture and engineering principles |
Organizations that hold ISO 27001 certification usually find that most CC6.1 controls already exist. The gap tends to be evidence over time: ISO 27001 audits sample on a certification cycle, while a SOC 2 Type 2 needs evidence that each control operated throughout the defined period.
How should you prepare for CC6.1? A six-step approach
- Reconcile your inventory. Compare the asset inventory, the system description, and what actually runs in your cloud accounts. Resolve every difference.
- Confirm SSO and MFA coverage. List every in-scope system and how users authenticate to it. Close or document gaps.
- Lock down the highest-privilege accounts. Root, owner, and break-glass accounts get hardware MFA, monitoring, and a written procedure.
- Inventory non-human credentials. Assign owners to service accounts and API keys, and revoke those nobody can explain.
- Verify encryption and key administration. Check encryption at rest on every data store and backup, and restrict who can administer keys.
- Make architecture reviews leave a record. Add a security sign-off to the design review template for new services.
If you would like an independent view before your auditor arrives, a SOC 2 readiness assessment from Ledger Audits tests your CC6 controls the way a service auditor would and gives you time to fix what it finds.
Related reading
- SOC 2 CC6.4: Physical Access Controls Across Cloud, Co-location, and Remote Setups
- SOC 2 Sample Sizes: How Auditors Decide How Many Items to Test
- Complementary User Entity Controls (CUECs) Explained
- SOC 2 System Descriptions and DC Section 200
- How to Verify a SOC 2 Report
- External: AICPA 2017 Trust Services Criteria (With Revised Points of Focus, 2022)
Frequently asked questions
What does SOC 2 CC6.1 require?
SOC 2 CC6.1 requires a service organization to implement logical access security software, infrastructure, and architectures over its protected information assets to protect them from security events. In practice, this means central identity and authentication, restricted access to systems and data, network segmentation, managed external access points, encryption at rest where risk warrants it, and protection of encryption keys.
How many points of focus does CC6.1 have?
CC6.1 has 13 points of focus in the 2017 Trust Services Criteria with the 2022 revised points of focus. Eleven apply to every SOC 2 examination. Two additional points of focus apply only when the Confidentiality or Privacy category is in scope. Points of focus are illustrative, not mandatory.
Is MFA required for SOC 2?
MFA is not explicitly required by the SOC 2 criteria. However, CC6.1 requires logical access security appropriate to your risks, and for SaaS companies an auditor will typically expect MFA for workforce access and for all access to production infrastructure. Missing MFA on privileged access is difficult to justify in a risk assessment.
Is encryption at rest required for SOC 2?
Encryption at rest is not a standalone SOC 2 requirement. The CC6.1 point of focus calls for encryption to supplement other protections where your risk assessment deems it appropriate. Because enabling encryption on managed cloud data stores is usually simple, most service organizations encrypt databases, storage, and backups and protect the keys through a managed key service.
What is the difference between CC6.1 and CC6.2?
CC6.1 covers the access control architecture itself: identity systems, authentication, network boundaries, and encryption. CC6.2 covers the user lifecycle: registering and authorizing users before issuing credentials, and removing credentials when access is no longer authorized. A misconfigured MFA policy is a CC6.1 issue. A terminated employee who still has an active account is a CC6.2 issue.
Does CC6.1 apply if we are fully hosted on AWS, Azure, or Google Cloud?
Yes. CC6.1 applies to the layers you control, including identity and access management, network rules, encryption settings, and key policies in your cloud accounts. The cloud provider's own controls are usually carved out of your report as a subservice organization, and you are expected to monitor them by reviewing the provider's SOC 2 report.
Get independent eyes on it before your auditor or buyer does.
Ledger Audits provides readiness and internal audit work only. We never issue SOC 2 reports or ISO certificates, which keeps our review independent of your examination.
Book a discovery call