Fingerprint scan


RPA and the government audit

Audit automation for increased compliance and cost efficiencies

As government programs evolve, and agency missions grow increasingly complex, governments are increasingly reliant upon technology to support strategic, operations, compliance, and reporting objectives. Faced with unprecedented disruption of the “old ways” of achieving these objectives, agencies have increased use of robotic process automation (RPA)—computer-coded, rules-based machines to automate repetitive, rules-based human tasks. In the simplest terms, RPA uses robots, or “bots,” to record human actions, then follow prescribed protocols and procedures with precision for increased compliance and cost efficiencies.

How is government using RPA?

​In its January 2018 blog post, “We’re thinking about robots,” the US Department of the Treasury’s Bureau of Fiscal Service (Fiscal Service) noted that the Office of Financial Innovation and Transformation (FIT) was tasked with piloting a project to explore ways RPA could improve financial processes in the bureau. The pilot discovered an average 60 percent improvement in the amount of time it took to finish all tasks in the seven processes that were automated. In addition, throughput increased by 30 times, and processing capacity was created without adding human resources. Because RPA performs tasks exactly as defined (assuming bots are properly configured), accuracy in carrying out automated tasks reached 100 percent.

Accounting is the most common area of RPA deployment in government, and this functional area of government is well-suited for many types of automation. For example, traditional RPA that automates transactional processes in accounting or enhanced finance and controls automation, often referred to as “the last mile of finance” automation, work well for a variety of reasons, including:

  • Need for a high degree of accuracy and consistency
  • Repetitive, manual nature of transaction processing
  • Information gathered from fragmented systems
  • Dependency on data entry, data manipulation, and report generation

Back to top

The changing regulatory landscape

With the increasing use of emerging technologies such as RPA, many regulatory bodies and standards setters have begun to assess them and form “regulator’s perspectives.” In prepared remarks given at the 43rd World Continuous Auditing and Reporting Symposium on November 2, 2018, Kathleen M. Hamm, a member of the Public Company Accounting Oversight Board (PCAOB), said innovative uses of technology offer the potential to fundamentally change not only the financial reporting process, but also the audit. She also said technology offers the promise of improving audit quality plus the opportunity to proactively develop accounting and auditing standards and regulatory requirements into solutions. Already, the newly revised Office of Management and Budget (OMB) Circular No. A-123, Appendix A—Management of Reporting and Data Integrity Risk, emphasizes data as an asset and requires more rigor related to compliance with the Digital Accountability and Transparency Act (DATA Act).

Moreover, at the American Institute of Certified Public Accountants (AICPA) conference on current developments at the Securities and Exchange Commission (SEC) and PCAOB, SEC officials stressed the importance of the Committee of Sponsoring Organizations of the Treadway Commission (COSO) Framework Principle 9 to maintain effective internal controls over financial reporting, particularly in a period of change like RPA implementation. Therefore, government financial managers should prepare for auditors to focus on the following areas in their financial statement audits:

  • Understanding the changes resulting from processes that have been automated and reevaluating risks through risk assessment
  • Technology reliability within business processes, including automated controls
  • Information system controls, including those for access and bot credentialing, change related to maintaining bots, business continuity planning, and other general controls
  • Accuracy, security, and completeness of stored data, including data leakage and privacy
  • Cybersecurity and procedures for cyber incident response
  • Segregation of duties between bot IDs and end users
  • RPA governance, including monitoring of bot throughput, processing errors and exception handling through human intervention (i.e., partial automation)
  • Completeness and accuracy of transactions processed through RPA, considering configuration, parameters of reports, and processing logic

Back to top

Touch screen

The auditors’ risk assessment

Auditors must seek to understand automated transactions in the entity’s operations that are significant to financial statements as well as related control activities. This can be conducted in walkthroughs of the automated processes and by understanding the related parameters, logic, and source data that the automation encompasses. It is important for auditors to understand those transactions that give rise or contribute to risks of material misstatement in financial statements. This includes a consideration of whether automation of transaction processing would result in modification of the nature, timing, and extent of auditors’ procedures. The information obtained from risk assessment would form the basis for the design of further audit procedures to respond directly to each risk.

Despite the evolution of RPA in accounting and finance, regulators remind auditors that the fundamental financial reporting framework does not change. If a significant amount of an entity’s financial information is electronically initiated, recorded, processed or reported, substantive audit procedures alone might not provide sufficient evidence for auditors for a relevant financial statement assertion. In other words, when RPA is used to meet financial reporting or internal controls requirements, auditors must comprehend the design of that technology. Effective internal controls are essential to mitigate the potential for improper initiation or alteration of information. These actions can go undetected when recorded, processed or reported in electronic form alone.

Regulators have cited challenges, such as technology that does not operate as intended due to coding errors when developed or intentional or unintentional changes made after the technology is deployed. In addition, in changing environments, computer code underlying complex technology can degrade over time, becoming less responsive. As a result, processes should be in place to continuously monitor and confirm that the output of an application remains consistent with expectations. Accordingly, audit firms have begun asking their clients specific questions during the planning phase of an audit, such as: “Have you implemented or do you plan to implement any emerging technologies to support financial reporting (for example, cloud computing, RPA or blockchain)?” This helps auditors gain a better understanding of RPA tools in the IT environment, which can impact the evaluation of the design of internal controls.

Back to top


Consideration of risks and internal controls

Government agency leaders should consider the revised OMB Circular A-123, particularly its requirements for enterprise risk management (ERM), which requires an increased focus on developing and utilizing a risk-based approach to achieve agencies’ strategic, operations, compliance, and reporting objectives. RPA solutions may introduce new risks to an organization if not managed appropriately. An increased focus on the effectiveness of an agency’s internal controls systems related to RPA informs how these systems help achieve objectives. The Government Accountability Office (GAO) Green Book sets standards for effective internal controls in federal agencies. Leaders should follow these standards when assessing RPA, building a business case for it, determining an optimal operating model, identifying RPA solutions, and planning implementation. They should also understand RPA through the lens of internal controls. In audit planning, auditors will seek to understand automated processes as well as internal controls relevant to the audit.

Improper implementation or automation of the wrong processes (e.g., operational risks) may result in immediate financial losses to an agency. Bot-related errors affecting the integrity of cybersecurity programs or compliance with data privacy regulations may not only result in direct costs to the agency but also cause reputational concerns. Therefore, it is critical to determine how changes inform risk assessment, particularly those risks arising from IT, and whether modifications to existing standards, processes, and structures (e.g., the control environment) are necessary. As agencies implement RPA, it is imperative they perform their own risk assessments associated with the processes being automated, design internal controls for it, and produce appropriate audit evidence.

Back to top

Person watching a screen

RPA governance in government agencies

​An effective governance model establishes accountability throughout the automation lifecycle, from the creation of the automation program to design and testing of bot functionality and outputs to bot implementation and, finally, to monitoring effectiveness of the automation program. It is important to identify an executive sponsor with appropriate knowledge of emerging technologies plus the authority to champion and lead the project in the agency. Oversight bodies may outline and develop a governance structure to encompass the following:

  • Establish RPA policies and standards through a governance framework
  • Leverage existing internal controls and enhance and adapt them to the automated environment
  • Define and create access management policies and internal controls for bots to oversee the work of other bots
  • Adapt bots to detect and report errors so that human intervention can occur when needed
  • Manage bots through environment changes by implementing appropriate controls
  • Create compliance policies and tools to monitor and log bot output for enhanced transparency
  • Establish cross-functional working groups and assign responsibility to individuals for maintaining bots and resolving issues.

RPA can be an effective solution for automating repetitive tasks, improving processes subject to audit, and contributing to effective risk management. RPA can help an organization achieve effectiveness and efficiencies in executing tasks that are critical to mission support, including timely reporting of financial and operational information and enhanced financial reporting. Bots not only follow prescribed protocols and procedures with precision, but also can be programmed to capture and maintain complete audit logs and automate reporting for a stronger audit process. Benefits for auditors include access to more audit data in a standardized, reliable, consistent format plus automated reporting that enables auditors to focus on analysis and decision-making, rather than manual data collection and consolidation.

Back to top