Understanding PCI-DSS

Requirement 1.2.2: Controlling Changes to Network Connections and Network Security Controls

Written by Derek Rogers | Nov 17, 2026, 4:13:44 PM

Requirement 1.2.2 governs how organizations change the devices and rules that control network traffic around payment card data. Every change to the connections between networks, and every change to the configuration of a firewall, router access list, or cloud security group that protects the payment environment, must pass through the same formal process the organization uses to approve and manage other production changes. The objective is to prevent a single unreviewed change from opening an insecure service or an unauthorized path to cardholder data.

To meet this requirement, organizations should confirm that their change procedure explicitly covers network changes, then apply it consistently. Each change should record the reason for the change and its security impact, receive approval from someone with the authority and knowledge to judge it, be tested before and after implementation, and include a way to return to a secure state if it fails. Once a change is verified, network diagrams and configuration standards should be updated so that documentation continues to match the environment.

During an assessment, assessors identify changes directly from network configuration settings and then trace each one back to a change record. A change that appears in a configuration but has no matching approved record is a direct finding, regardless of how well the written procedure reads. 

 

Requirement Information

Parent Requirement 1.2

Network security controls (NSCs) are configured and maintained.

PCI DSS Requirement 1.2.2

All changes to network connections and to configurations of NSCs are approved and managed in accordance with the change control process defined at Requirement 6.5.1.

Applicability Notes

Changes to network connections include the addition, removal, or modification of a connection.

Changes to NSC configurations include those related to the component itself as well as those affecting how it performs its security function.

Priority: 6

Priority Meaning: Complete remaining compliance efforts, and ensure all controls are in place. This milestone completes PCI DSS requirements and finishes all remaining related policies, procedures, and processes needed to protect the cardholder data environment.

Summary of Control Requirements

Requirement 1.2.2 connects network security to the organization's broader change management program. It does not ask network teams to build a separate process. Instead, it requires that network changes follow the change control process defined at Requirement 6.5.1, with the same documentation, approval, testing, and recovery elements that apply to any other production change.

Definition: Change Control

Change control is the documented process an organization uses to request, evaluate, approve, implement, verify, and record changes to production systems.

For example, a network engineer opens a ticket describing a new firewall rule, a manager with authority over the payment environment approves it, and the engineer attaches the post-change test results before the ticket is closed.

The requirement covers two categories of change. The first is changes to network connections, meaning any connection that is added, removed, or modified. The second is changes to the configuration of network security controls, including changes to the device or service itself and changes that affect how it performs its security function.

Definition: Network Connection

A network connection is any logical or physical path that allows traffic to flow between two systems or networks.

For example, adding an internet service link at a retail location, creating a site-to-site tunnel for a payment processor, or peering a new cloud network with the network that hosts the payment application are all changes to network connections.

Definition: Network Security Control (NSC)

A network security control is a technology that manages, filters, inspects, or restricts network traffic based on configured rules.

For example, an on-premises firewall, a router configured with an access control list, and a cloud security group that limits which systems can reach a payment application are all network security controls. 

 

Why This Requirement Exists

PCI DSS states the purpose of this requirement directly: "Following a structured change control process for all changes to NSCs reduces the risk that a change could introduce a security vulnerability."

Network security controls are only as reliable as the last change made to them. A single rule added in a hurry can expose a database to the internet, allow an insecure protocol into the payment environment, or bypass segmentation that the organization relies on to limit its PCI DSS scope. A new connection, such as a site-to-site tunnel to a vendor or a peering link between cloud networks, can create a path into the cardholder data environment that does not appear on any diagram.

These exposures rarely come from bad intent. They usually come from ordinary work: a temporary rule for a vendor that is never removed, an emergency fix during an outage that is never reviewed, or a configuration change that quietly disables logging or inspection. Without a structured process, no one evaluates the security impact before the change, and no record exists afterward to show who approved it or why.

Requirement 1.2.2 addresses both the security and governance sides of this problem. Approval establishes accountability for each change. Testing confirms that the change performs as expected without weakening security. Documentation keeps network diagrams, configuration standards, and periodic rule reviews aligned with the environment that actually exists. 

 

Applicability of the Self-Assessment Questionnaire (SAQ)

PCI DSS Requirement 1.2.2 appears in the following Self-Assessment Questionnaires:

  • SAQ A-EP: E-commerce merchants that partially outsource payment processing and whose website does not receive account data but controls how customers, or their account data, are redirected to a PCI DSS compliant third-party service provider.
  • SAQ D (Merchant): Merchants that are eligible to complete a self-assessment questionnaire but do not meet the criteria for any other SAQ type.
  • SAQ D (Service Provider): Service providers defined by a payment brand as eligible to complete a self-assessment questionnaire.

 

Scope

Requirement 1.2.2 applies to every network security control that protects the cardholder data environment or controls traffic between the cardholder data environment and other networks. It also applies to every network connection into, out of, or within that environment. This includes on-premises firewalls, router access control lists, cloud security groups and virtual network rules, and VPN gateways that terminate connections reaching in-scope systems.

Scope does not end where an outside provider manages the device. When a third party administers an organization's firewall or cloud network, changes made by that provider must still be approved and managed in accordance with Requirement 6.5.1, and the division of that responsibility should be documented.

PCI DSS requirements apply to:

  • The cardholder data environment (CDE), which is comprised of:
    • System components, people, and processes that store, process, or transmit cardholder data and/or sensitive authentication data, and,
    • System components that may not store, process, or transmit CHD/SAD but have unrestricted connectivity to system components that store, process, or transmit CHD/SAD.

AND

  • System components, people, and processes that could impact the security of cardholder data and/or sensitive authentication data.

Definition: VPN Gateway

A VPN (virtual private network) gateway is a device or cloud service that creates encrypted tunnels between networks or for remote users, and it often enforces rules about which traffic may pass through those tunnels.

For example, a VPN gateway that lets a support vendor reach a payment server is in scope, and adding a new tunnel or widening the systems that vendor can reach is a change that must follow Requirement 1.2.2. 

 

Understanding the Requirement

Requirement 1.2.2 contains four practical expectations.

Network Changes Are Included in the Formal Change Control Process

The documented change control procedure must state that it applies to network connections and to NSC configurations. A general procedure written for servers and applications is not sufficient if network changes are handled informally by a separate team or vendor.

Each Change Contains the Elements of Requirement 6.5.1

Requirement 1.2.2 relies on the elements defined at Requirement 6.5.1. PCI DSS Requirement 6.5.1 reads:

Changes to all system components in the production environment are made according to established procedures that include:

  • Reason for, and description of, the change.
  • Documentation of security impact.
  • Documented change approval by authorized parties.
  • Testing to verify that the change does not adversely impact system security.
  • For bespoke and custom software changes, all updates are tested for compliance with Requirement 6.2.4 before being deployed into production.
  • Procedures to address failures and return to a secure state.

Applied to network changes, this means each change record should describe what is changing and why, state how the change affects security (for example, which systems gain or lose access), show approval by an authorized person, include evidence of testing, and identify how the change will be reversed if it fails.

Changes Are Approved Before They Are Implemented

PCI DSS guidance states that changes should be approved by individuals with the appropriate authority and knowledge to understand the impact of the change, and that all changes should be approved prior to being implemented and verified after they are implemented. Organizations that need an expedited path for outages should define that path in the procedure, including who may approve it, rather than allowing emergency work to bypass approval.

Definition: Emergency Change

An emergency change is a change that must be made faster than the normal approval cycle allows in order to restore service or address an urgent security issue.

For example, if a firewall rule is blocking payment authorizations during peak hours, an organization may use a predefined emergency path in which an on-call manager approves the fix by phone and the full record is completed and reviewed the next business day.

Network Documentation Is Updated After Each Change

PCI DSS guidance states that once a change is approved and verified, network documentation should be updated to prevent inconsistencies between network documentation and the actual configuration. This keeps network diagrams, data-flow diagrams, and configuration standards accurate, and it supports the periodic NSC configuration reviews required elsewhere in Requirement 1. 

 

Additional Technical Details

The following practices help IT teams produce reliable evidence for Requirement 1.2.2 without adding unnecessary process.

Detecting Changes Independently of Change Records

Assessors identify changes from configuration settings, not from the change log. Internal teams should be able to do the same. Common methods include:

  • Scheduled configuration backups with automated comparison between versions
  • Configuration change events forwarded from NSCs to a central log
  • Cloud provider audit logs that record security group and network changes, such as AWS CloudTrail or Azure Activity Log
  • Revision history in a firewall management console

Linking Configurations to Change Records

Many firewalls and cloud platforms allow a comment, description, or tag on each rule. Recording the change ticket number in that field allows any reviewer to trace a rule to its approval in seconds. It also makes stale or undocumented rules easy to spot during six-month NSC reviews.

Infrastructure as Code

Organizations that manage cloud networks through code can use their code repository as part of change control. A pull request can carry the description, security impact, approval, and automated test results, provided the approval comes from an authorized party and the repository is the only path to production.

Definition: Infrastructure as Code (IaC)

Infrastructure as code is the practice of defining network and system configurations in version-controlled files that are applied automatically, rather than configuring each component by hand.

For example, a cloud security group rule written in a template file is reviewed and approved in a pull request, then deployed by a pipeline that rejects changes made outside that workflow.

Pre-Change and Post-Change Verification

Testing for network changes can be simple and still meaningful. Examples include confirming intended connectivity, confirming that traffic which should remain blocked is still blocked, comparing the new configuration against the configuration standard from Requirement 1.2.1, and confirming that logging remains enabled on the changed rule or device. 

 

Ways to Meet Requirement 1.2.2

Requirement 1.2.2 may be satisfied through several approaches, depending on the size of the environment and the tools already in use.

Option 1: Extend the Existing IT Change Process

Add a network change category to the change procedure already used for servers and applications. Include required fields for affected connections or rules, security impact, approver, test results, and rollback steps. This is often the lowest-cost approach because it reuses an existing process and approval structure.

Definition: Rollback Plan

A rollback plan is a documented set of steps for reversing a change and returning a system to its previous secure state if the change fails or causes problems.

For example, before modifying a firewall policy, an administrator saves the current configuration and records the exact command or restore step needed to reapply it within minutes if payment traffic is disrupted. 

Option 2: Use a Standard Network Change Template

For smaller teams, a short template in an existing ticketing or help desk system can capture every element of Requirement 6.5.1. Pre-approved standard changes, such as routine low-risk updates with a documented procedure, can reduce approval overhead while keeping each change recorded.

Option 3: Manage Cloud NSCs Through Code Pipelines

Where cloud security groups and network rules are defined as code, require that every change pass through a reviewed and approved pull request, and restrict direct console changes to documented emergency use.

Option 4: Reconcile Changes Against Records

On a defined schedule, compare configuration differences against closed change tickets. Any difference without a matching record is investigated and either documented retroactively as an exception or reversed.

Option 5: Define Change Responsibilities With Third Parties

When a provider manages NSCs, document how the provider requests, approves, and records changes, how the organization is notified, and who approves changes that affect the cardholder data environment. Request change records from the provider as part of routine oversight. 

 

Compensating Controls and Customized Approaches

Customized Approach Objective

PCI DSS defines the customized approach objective as:

Changes to network connections and NSCs cannot result in misconfiguration, implementation of insecure services, or unauthorized network connections.

An organization using the customized approach must demonstrate that its method prevents these outcomes as effectively as the defined approach. The customized approach cannot be documented in an SAQ. Entities that wish to use it should consult the organizations that manage their compliance program, such as their acquirer or payment brands.

Practical Customized Approach Examples

Potential approaches include:

  • Policy-as-code controls that automatically block network changes violating the configuration standard
  • Continuous configuration monitoring that detects and reverts unapproved changes
  • Network management platforms that permit changes only through an approval workflow

Each approach must still be supported by evidence that unauthorized or insecure changes cannot persist in the environment.

Compensating Controls

Compensating controls are rarely necessary for Requirement 1.2.2 because change control can be implemented with modest tools. Where a legitimate and documented constraint exists, such as a legacy device that cannot export configuration history, a compensating control might combine restricted administrative access, manual configuration snapshots, and frequent independent comparison. Any compensating control must address the objective of the original requirement and be documented in a Compensating Controls Worksheet.

SMB Considerations

Small and midsized organizations do not need a change advisory board or a dedicated firewall management platform to meet this requirement. A clear written procedure, a ticket template that captures each element of Requirement 6.5.1, a named approver, and a periodic comparison of configuration backups against tickets are generally sufficient for a small number of NSCs. 

 

Demonstrating Compliance

Organizations must be able to show that network changes are covered by the change process and that individual changes followed it.

Baseline Evidence Requirements

Examples include:

  • Documented change control procedure that explicitly includes network connections and NSC configurations
  • Change records for network changes within the assessment period
  • Approval records identifying the authorized approver
  • Documented security impact for each change
  • Pre-change and post-change test results
  • Rollback or failure procedures
  • Current network configurations and updated network diagrams

Enhanced Evidence for Maturity

Organizations with higher maturity typically maintain:

  • Automated configuration comparison reports reconciled against change tickets
  • Change ticket references recorded in rule comments or tags
  • Trend reporting on emergency and unauthorized changes
  • Management review of change exceptions
  • Documented change responsibilities for third-party managed NSCs
  • Historical records showing the process operating consistently across the year

Frequently Missing Evidence

Assessors frequently request evidence that organizations cannot produce, including:

  • Approval dated before implementation
  • Records for changes made by a managed service provider or cloud administrator
  • Documentation of the security impact, rather than a description of the change alone
  • Test results showing the change was verified
  • Follow-up review of emergency changes
  • Evidence that network diagrams were updated after a connection changed

 

Testing Procedures

PCI DSS Testing Procedure 1.2.2.a

Examine documented procedures to verify that changes to network connections and configurations of NSCs are included in the formal change control process in accordance with Requirement 6.5.1.

Application of the Testing Procedure by Internal Teams

Internal Audit, Compliance, and Information Security should read the change control procedure as an assessor would and confirm that network changes are clearly covered.

Key review questions include:

  • Does the procedure explicitly name network connections and NSC configurations?
  • Does it require every element of Requirement 6.5.1 for network changes?
  • Does it cover cloud network rules and changes made by third parties?
  • Does it define how emergency changes are approved and reviewed?
  • Does it require network documentation to be updated after a change?

Relevant examples include:

  • Change management policy and procedures
  • Network change ticket templates
  • Emergency change procedures
  • Third-party responsibility documentation

Typical findings include:

  • Procedures that cover servers and applications but not network devices
  • No mention of cloud security groups or virtual network rules
  • Templates missing security impact or rollback fields
  • Undefined emergency change handling

PCI DSS Testing Procedure 1.2.2.b

Examine network configuration settings to identify changes made to network connections. Interview responsible personnel and examine change control records to verify that identified changes to network connections were approved and managed in accordance with Requirement 6.5.1.

Application of the Testing Procedure by Internal Teams

Internal teams should start from the environment, not the ticket queue. Identify connections that were added, removed, or modified during the period, then trace each one to a change record.

Key review questions include:

  • Which VPN tunnels, circuits, cloud peering links, or interfaces changed during the period?
  • Does each change have a record with reason, security impact, approval, testing, and rollback steps?
  • Was approval granted before implementation?
  • Do network diagrams reflect the current connections?

Relevant examples include:

  • Configuration comparisons or revision histories
  • Cloud audit logs showing network changes
  • Change tickets for each identified connection change
  • Current network diagrams
  • Interview notes from network administrators

Typical findings include:

  • Vendor or remote-access connections added without a change record
  • Decommissioned connections still present in configurations
  • Approval recorded after the change was implemented
  • Network diagrams that omit recently added connections

PCI DSS Testing Procedure 1.2.2.c

Examine network configuration settings to identify changes made to configurations of NSCs. Interview responsible personnel and examine change control records to verify that identified changes to configurations of NSCs were approved and managed in accordance with Requirement 6.5.1.

Application of the Testing Procedure by Internal Teams

Internal teams should sample NSC configuration changes from the period, including rule changes and changes to the device or service itself, and confirm that each followed the change process.

Key review questions include:

  • Were firewall rules, access control lists, or security groups modified without a ticket?
  • Were firmware updates, logging settings, or inspection features changed, and were those changes recorded?
  • Does each rule reference the change that created it?
  • Were temporary rules removed on schedule?

Relevant examples include:

  • Firewall and router configuration revision history
  • Cloud security group change logs
  • Change tickets with approvals and test results
  • Rule comments or tags containing ticket references
  • Records of emergency change reviews

Typical findings include:

  • Rules created directly in the console with no record
  • Temporary rules that became permanent
  • Device-level changes, such as disabled logging, treated as outside change control
  • Approvals by individuals without authority over the affected environment

 

Common Mistakes

Covering Rules but Not Connections

Organizations often route firewall rule changes through change control but add VPN tunnels, circuits, or cloud peering connections informally. Requirement 1.2.2 covers both.

Treating Device Changes as Maintenance

Firmware updates and changes to logging, inspection, or management settings affect how an NSC performs its security function. The Applicability Notes include these changes within the requirement.

Approving After the Fact

A ticket that is created and approved after the work is done records the change but does not control it. Assessors look at approval timing.

Leaving Emergency Changes Unreviewed

Emergency paths are reasonable, but changes made under them still need complete records and a follow-up review. Unreviewed emergency changes are a common source of permissive rules that remain indefinitely.

Letting Documentation Fall Behind

When diagrams and configuration standards are not updated after changes, the organization loses an accurate picture of its scope, and later reviews start from incorrect information.

Overbuilding the Process

Requiring formal board meetings for every rule change or purchasing specialized platforms before the basic process works adds cost without improving compliance. A consistent, well-documented ticket process is generally the more effective starting point for smaller environments.