SOX 404 & IT Controls: IT Control Recommendations For Small and Mid-Size Companies by
SOX 404 & IT Controls: IT Control Recommendations For Small and Mid-Size Companies by
SOX 404 & IT Controls: IT Control Recommendations For Small and Mid-Size Companies by
IT Control Recommendations For Small and Mid-size companies by Ike Ugochuku, CIA, CISA
TLK Enterprise 2006, www.tlkenterprise.com
INTRODUCTION
Small, medium, and large businesses today face similar challenges concerning assuring the public that their published financial results actually reflect their performance. The increasing reliance on technology for all business processes and transactions has led to greater technology expense and a need for a controlled technology structure to ensure that transactions are reliable and accurate. This gives senior management a greater assurance that the companys financial results actually represent the performance and that the organization is compliant with regulatory requirements. Recently, the Committee of Sponsoring Organization (COSO) released a document Guidance for smaller public companies reporting on internal controls for financial reporting. The document is to assist smaller companies that desire better controls for their business. The question of how to translate these recommendations into practical application is covered by this document. This document gives practical advice concerning controls for IT operations.
Source
It is based on the authors years of experience working with large, mid-size, and small corporations. The author is an established technology risk audit professional experienced in analyzing IT and business processes, creating and implementing corrective recommendations to mitigate identified risks or gaps. He has been involved in the development and implementation of IT policies for several firms and has an experiential knowledge of truly effective policies that create a controlled IT environment.
Audience
This document is high level enough for business senior management who want to understand about IT processes. It is also suitable for IT management who want to understand about control requirements for IT processes. IT auditors and compliance officers will benefit from the information in this document.
OVERVIEW
The need for controls
It is increasingly important to have an IT department with proper internal controls built into its operations. Two major factors are causing senior management to take a closer look at the IT controls in their organizations. Regulatory Compliance: new laws like Sarbanes-Oxley (SOX) specify that select senior officers will attest to the state of the internal controls of the company and the reliability of the financial statement released. Since almost all company transactions involve use of a computer system, it is important that there are adequate systems controls in place to assure regulators that the attestation by management is reliable. IT expenditure: it is also important that IT expenditure is aligned with the companys business strategy. As the reliance on technology systems to perform and report on transactions increases, technology cost is one significant internal expense. It is important that this cost is aligned with business strategy. There should be a documented IT process that integrates business participation in the creation of an IT strategy plan. The short-term approach to verify adequate internal controls is to test existing controls and recommend compensating or new controls for weak controls. A long term approach is to build controls into the IT planning processes. This can be achieved by using a standard IT framework like Control Objectives for Information and related Technology (COBIT) at the high level. But for implementation it may be necessary to hire expensive consultants to map the processes to the COBIT framework. For small and mid-size companies with IT, a better, more cost effective way is to have a comprehensive information systems policy document (like this document) which can be used as a control guideline for assessing current operations and creating new processes. It also serves as a baseline that can be used by auditors to review if the IT departments operations are consistent with business objectives. This document is based on the COBIT framework. It uses the 12 control objectives that are closely aligned with the US Public Accounting Oversight Board (PCAOB) general control guidelines for IT systems. It provides recommended controls that can be used to achieve these objectives.
effectiveness of the internal controls that govern processes that affect financial reporting. There should be an annual control assessment report with attestation from senior management. The US Public Accounting Oversight Board (PCAOB) was creating by the SOX act. In order to protect the interest of investors, the board is to oversee the audit of public companies. The PCAOB auditing standard No.2 discusses the audit of internal controls that govern financial reporting. It states that since most financial transactions involve IT systems, the controls for IT operations should be examined. The standard identifies four key IT processes, namely, program development, program changes, computer operations, and access to programs and data. The Control Objectives for Information and related Technology (COBIT) is a set of best practices (framework) for information (IT) management created by the Information Systems Audit and Control Association (ISACA), and the IT Governance Institute (ITGI). Control Objectives for Information and related Technology provides managers, auditors, and IT users with a set of generally accepted measures, indicators, processes, and best practices to assist them in maximizing the benefits derived through the use of information technology and developing appropriate IT governance and control in a company. In its fourth edition, COBIT has 34 high level objectives that cover 215 control objectives categorized in four domains: Plan and Organize, Acquire and Implement, Deliver and Support, and Monitor and Evaluate. This document is based on IT control objectives for Sarbanes-Oxley a document by ISACA, which maps 12 of the COBIT control objectives to the four key IT processes identified by the PCAOB. The chapters discuss relevant policies and procedures that can be used to achieve the 12 objectives. The information provides a logical starting point for affected organizations.
COBIT Control Objective heading 1. Acquire and develop application software 2. Acquire technology infrastructure 3. Develop and maintain policies and procedures 4. Install and test application software and technology infrastructure 5. Manage changes 6. Define and manage service levels 7. Manage third-party services 8. Ensure systems security 9. Manage the configuration 10. Manage problems and incidents 11. Manage data 12. Manage operations
Development
Ooperations
Computer
Program
Program
Changes
DATA MANAGEMENT
Data management and control depend on what the data is used for and the classification of the data. Proper classification of data assists management in making decisions about data protection expenses.
Data ownership
The data owner should be a business person or user. A business person refers to a unit that has no direct management of the infrastructure that the data resides on. The user will determine who should have access rights to the data; authorize any changes to the data, determine appropriate backup and recovery times and be responsible for any cost related to maintenance of the data.
Data classification
Data should be classified to ensure that the right kind of security is implemented. The possible categories are the following: 1. Restricted: Data have a highly monitored distribution list. Where financially possible, data encryption should be enforced. Since encryption may be an expensive option for small size businesses, it may be useful to protect certain data with attorney-client privilege. 2. Confidential: Data are limited to only those defined as data owners. Data should be on an infrastructure that has a regular backup schedule. 3. Internal: Data are restricted to only employees. Data should have a regular backup schedule. 4. Public: Data that is already available to the public.
The User id
Because access controls are built around user accounts, the creation and management of user accounts is vital to an organization,. The purpose of user accounts is to grant employees access to specific network systems and resources. The user should only have one user id and password, which should be used for access to all systems on the network. A central authentication system may be established for applications so that developers dont have to concern themselves with how to get security information from the operating system. There may be several operating systems in the organization. The following controls should apply to user ids: 1. The user id should be related to personal information of the employee. This makes it easier for the employee to remember. It also makes it easy for the employee to be identified on the network. 2. The employee number should be used for uniqueness. Some small and mid-size companies may not have employee numbers, but where they have, a combination of employee number and initials can be used. This makes it difficult for unauthorized users to guess. 3. Three logon attempts should be allowed, after which the account is locked. It can only be unlocked after a formal request from the manager. The system should log the number of failed attempts for a user id, and this should be reviewed at least bi-annually. A report should be sent to responsible business managers of user ids with high failed attempts. 4. If the user id is not used for more than 90 days, it represents a security risk, so the user id should be disabled.
The password
Most small and mid-size companies may not assign employees a unique number, so user ids may be based on their names only. This makes it relatively easy to guess so it is important to have strong controls for the passwords. The following controls should apply to passwords: 6
1. At least 6 characters (numbers and alphanumeric) should be required. 2. There should be at least one number required. 3. There should be at least one uppercase letter required. 4. The user should be required to change the password every 60 days. 5. The system should keep a history of 8 passwords. These rules should be implemented as policy on the operating system. If a separate authentication system is used for some application, the rules should also be part of a system policy.
Account deletion
As soon as the employee contract is terminated, it is important that an account be disabled. This prevents the unauthorized account from being used to access resources. Human resources should notify IT of the exact details of an employees disengagement. Except in the case where there is a sudden termination, IT should be notified at least a week before the employees final date. Before being deleted, the account should be disabled for 30 days. Any exceptions to this should be submitted as a request from Human Resources.
Entitlement review
When users are moved within the company or if a project has ended, the resources rights of a user may change. It is important to have a process where the access rights are reviewed by the resource owners. All access rights to resources should be documented in a periodic report. The report should be prepared by the resource administrators or the information security officer. The Access report should show: 1. All the user ids that have accessed the resource within the period. 2. The names of the associated users 3. A list of resources that the users have accessed. 4. The access rights of that the users have to the resources. 5. Show the date of the last time that the user accessed the resource. The resource owners should review the entitlement report every 6 months. The company data security officer should coordinate the review process and make sure that the resource owners sign the review.
Remote access
Remote access extends the boundaries of the network and allows users to access company resources from outside the office. Remote access should be through a remote server which can allow the creation of user profiles. There should be at least two types of profiles. One, a normal user profile which will be used by a user that doesnt perform system administrative work. Two, an IT support user profile for users that perform system support work. The request for remote access should come from the manager; it should specify the kind of access needed and specify the primary location where access would occur. Connection The following guidelines are recommended for securing a remote access connection: 1. The remote access connection should have encryption for all connections. 2. The remote access connection should have a time out for all connections. 3. The remote access connection should have a log for all connections. 4. User id should be different from the network id. 5. The password should change with every connection. Small businesses that have their IT infrastructure hosted by an external vendor should ensure these controls exist in the vendors operations. If they have their own IT infrastructure, cost consideration should not prevent the installation of a remote system with good controls.
2. There should be a defined process to test and implement patches and updates. There should be a template test plan document. Implementation should follow the management process for change. 3. To verify system invulnerability, especially if an external website is available to the public, once a year an ethical hack should be conducted by an outside source. 4. Before being placed in production, applications should undergo an ethical hack
10
MANAGE OPERATIONS
Segregation of duty
For small and medium size companies, there are IT functions that should not be combined. Due to the size of the companies envisioned in the scope of this document, this could prove very difficult to enforce. The IT headcount should be proportionate to revenues generation and other financial factors. In situations where there is inadequate segregation of duties, the following recommendations are suggested and should serve as sufficient compensating controls: There should be an audit trail of all operations. Change should be as localized as possible. Change access rights should be as specific as possible. Developers should be separate from support staff. Where developers play a dual role of support, all changes should follow a change management process. 5. There should be regular independent review of the access logs. 6. There should be a maker-checker control on data entry operations. 1. 2. 3. 4.
the computer system, documentation should be reviewed and updated. This document change process should be part of the change control process. The architecture and design document should show the objective, business requirement, testing, technical specifications, business process integration and implementation strategy Alternate Hardware There should be alternate hardware stored in a different location. A hot site is a location that has all the required IT hardware and software installed and running. Because it may be too expensive for small and mid-size companies, there is no recommendation for a dormant hot site. Excess capacity may be built into the network design so that the COB site can be used for some production traffic. Management would have to consider cost considerations and required recovery speed. A cold site is a location where the company has access to IT hardware. For small or mid-size businesses, a lease contract with an external data center should be available. Requirements for alternate location would depend on overall business recovery strategy, available datacenters, and the number of employees at each location. At a minimum, mid-size companies should have an alternative cold site, while small companies should have a good backup strategy. System Backup and recovery There should be a backup schedule. This would depend on the amount of data in use and the recovery Service Level Agreement (SLA) with the business. The backup should be stored offsite periodically. This offsite storage should be updated daily if possible and weekly at a minimum. To ensure data integrity and to confirm the expected time to recover the data, backup data should be periodically restored. This process should be documented and reviewed by the data owners.
12
Problem reporting
13
System issues can be reported by the users or by the system administrators. There should be a formal process to create a support ticket to track the problem. A helpdesk ticket system can be purchased to make it more efficient. The following should be included: 1. Access specifications or profiles, 2. Ability to build policies into the system. (i.e., a request for a new employee id can only come from Human Resources), 3. Archiving of tickets, 4. Search feature. Small and mid-size organizations may outsource their helpdesk. It is important that there is a good workflow link with the vendors system. For instance, a request for creation of a network account for a new employee should go from Human Resources to the helpdesk. Or if an account is locked, the request to unlock the account should only come from the manager of the user.
14
MANAGE CHANGES
Change management is the process used to plan, schedule, apply, distribute, and track changes for all corporate activities. There should be a formal documented process for implementation of change. To help identify problems that may not be seen during the development phase, many organizations establish a laboratory to conduct tests before system changes are implemented. A lab should be set up where all testing is done. Since the lab is not a secure environment, test data should be used and not production data. If possible, the lab should be on a separate network segment. The staging area is a segment of the network set apart for tests to determine how proposed system changes would impact the production system. As much as possible, the staging environment should simulate the production. All changes to production should first be implemented in the staging. Specific users should be given access to the staging environment for User Acceptance Testing (UAT) testing. The staging environment should be secured and can not be on the same segment as production. A process for communicating change should be developed. The IT department will develop a document specifying the description of the change, the impact analysis, the system documents that need to be updated, the responsible support engineer, the time for the change and the fallback procedure. There should be a change committee which is comprised of at least one technology person and at least one person from the business. The committee will meet weekly to review all change requests. Once they have been approved, an announcement is made to the affected users. If there is an emergency change, the following apply: 1. The technology manager should approve the change, 2. There should be proper communication to affected parties, 3. After the emergency, the change form should be filled and approved, 4. There should be a post-mortem report on the change that should be sent to the business and technology manager(s).
Fall back
If there is a problem with the change, there should be a back to reverse the change and go back to the previous state. If there is a need for a fall back, there should be a review in the change control meeting to discuss reasons.
15
APPLICATION DEVELOPMENT
There are several software development models that a company can take; each depends on the project and the business requirements. For consistency reasons, it is recommended that the company adopt one method. The model chosen should involve the business in the beginning and end stages of development and require proper documentation at all levels.
Pre-implementation review
During the various stages of the SDLC process, the business security officer should review the application for adequate controls. If possible, an auditor should be involved in all stages of development, but this is not always possible in small and mid-size companies where the IT auditors are probably less than 5.
Internal SLA
For small businesses, an internal SLA may not be needed. However, even though it might change the informal culture, small businesses with IT departments of 20 or less should consider implementing an SLA within the business. Where an SLA doesnt exist, there should be a documented process for IT operations. The document should state details of the process. Information on the following items should be included in the document: 1. The expected delivery times to the business, 2. The expected result for the business. The document should be developed with the business. For mid-size companies, there should be an internal service level agreement between technology and other departments. The agreement should define the following items: 1. Expected service such as timeliness and accuracy of data, 2. Service thresholds for escalating a problem to higher management levels 3. The call tree for escalation. The SLA should be signed by the data owners.
External SLA
There should be an SLA with the vendor. The agreement is similar to the internal SLA defined above. It should state the following: 1. Expected service such as timeliness and accuracy of data, 2. Service thresholds for escalation, 3. The call tree for escalation.
Monitoring SLA
There should be a periodic report that states whether service level expectations were met. This report should be reviewed and signed by the data owners. The SLA should be for a period, for example, for 1 or 2 years. At the end of the period, the SLA should be reviewed and renewed or not.
17
IT strategy definition
For mid size companies, there should be an IT governance/steering committee consisting of business people and the technology manager. This committee reviews policies and defines IT strategy for the year. At the end of the second quarter, the steering committee meets to develop a formal IT strategic plan for the next year. In small businesses, it should be developed by the IT manager, who should be part of the business strategy discussions for the year. Since this determines the budget for technology, the data owners should sign an approval of the IT strategic plan.
Capacity planning
Based upon current resource utilization, historical trends, and projected business needs, this is the annual forecasting of future hardware, software, and network resource requirements. Every three months a review should be done by IT management of current resource utilization to detect notable trends. Showing performance and availability metrics of the network system, a periodic (preferably every quarter) report should be sent out. The report should show comparison with historic and expected figures. The report can contain items such as the memory and CPU utilization, server uptime and average user response. The steering committee should discuss this report and, based on expected business needs, decide if the current or projected resources are adequate. At the end of the second quarter of the year, the steering committee should meet to decide resource requirements for the coming year. This should be submitted to the company budget officer by the end of the third quarter.
Purchase justification
Each acquisition should have a business justification related to either defined IT strategy or capacity planning. Unplanned requests should be justified by showing evidence of related support issues or new project request(s) from the business.
18
19
20