审计任务自动化：机器人流程自动化评估框架

需积分: 10 197 浏览量更新于2024-07-09 收藏 526KB PDF 举报

"使用机器人流程自动化执行审计任务的框架-研究论文" 这篇研究论文探讨了在审计领域如何有效地利用机器人流程自动化（RPA）技术，特别是对于审计任务自动化优先级的确定。随着科技的进步，RPA，即机器人或自动化的应用，已经显著改变了业务运营方式。然而，审计人员在决定哪些任务适合自动化以及如何优先考虑这些任务时面临挑战。研究者通过设计科学方法建立了一个三步评估框架，以帮助内部和外部审计师在RPA实施过程中做出关键决策。这个框架旨在解决审计行业的痛点，协助审计员更高效地运用RPA技术。框架的验证过程包括了对有经验的内部和外部审计师的访谈、调查，以及两个实际案例的研究，这表明框架对于提高审计过程中的数字化劳动力（digital labor）和数字审计（digital auditing）具有积极作用。论文中提到的三步评估框架可能包括以下步骤： 1. 任务识别：首先，审计师需要识别出适合自动化的工作任务。这涉及到对现有审计流程的深入理解，找出那些重复性高、规则明确、易出错的任务。 2. 优先级设定：一旦识别出可自动化任务，就需要根据任务的重要性和紧迫性来设定优先级。这可能基于任务的频率、复杂性、潜在的错误风险和自动化带来的效率提升。 3. 实施与优化：最后一步是实施RPA解决方案，并持续监控和优化。这包括选择合适的RPA工具，创建和测试机器人，以及在实际操作中不断调整以提升效果。这篇研究为未来会计领域的研究者提供了一个模型，他们可以参考这种方法论来开发针对新兴技术的创新框架。同时，论文也强调了RPA在审计领域的潜力，不仅可以提升工作效率，还能减少人为错误，使审计工作更加精确和可靠。总结而言，这篇论文的核心贡献在于提供了一个实用的评估框架，它可以帮助审计行业更好地利用RPA技术，推动审计工作的数字化转型。通过这个框架，审计团队可以更有策略地引入RPA，从而提高审计质量，降低风险，同时释放审计人员的时间，让他们能专注于更高价值的分析和决策支持工作。

展开



Big 4 Accounting Firms

We inquired with the Big 4 accounting firms about internal frameworks they may

have developed to help auditors or clients automate tasks. For each Big 4 accounting firm, we

approached the following individuals (we list titles and areas of focus):

 Firm 1: Director in National Office (assurance), Partner (risk management and

internal audit), Principal (risk management), Associate Partner for Intelligent

Automation (Consulting Area)

 Firm 2: Senior Manager (assurance), Partner (internal audit)

 Firm 3: Managing Director (assurance), Managing Director (risk and financial

advisory services), Director (risk advisory services), Senior Manager (risk advisory,

focused on digital tools including RPA), Senior Manager (assurance and risk

advisory)

 Firm 4: Partner (internal auditing), Senior Partner and Practice Head (internal auditing

and compliance management), Director focusing on Digital (assurance), Senior

(advisory group focused on building bots for clients), Senior (risk assurance, focused

on digital tools including RPA).

We had an open discussion with each individual about selecting tasks for automation.

For individuals in the auditing space, we asked how tasks were chosen for automation. For

individuals in the consulting space, we asked how they helped clients choose tasks for

automation (we also inquired about if and how they selected tasks for automation within the

firm). While we started out broad to understand their perspective, we also directly asked if

their firm had firm-wide frameworks for deciding which tasks to automate or if the firm had

approved frameworks to use with their clients. All individuals were unaware of any firm-

wide frameworks. When we asked how employees decided what tasks to automate or how

they advised clients on what to automate, they said it was largely done based on past

experience and individual judgment and that it varied significantly from individual to

individual or from team to team. We finally asked about the need for having a framework to



We received appropriate university permissions to use human participants in this study.

Electronic copy available at: https://ssrn.com/abstract=3651028



help decide what tasks to automate. The individuals we interviewed agreed that a validated

framework for helping auditors or clients know what tasks to automate would be beneficial.

Our inquiries were directed towards auditors at the Big 4 accounting firms. Given that

smaller public accounting firms are likely to have fewer resources, we would expect that the

need for a framework is even greater for these firms.

Survey of Internal Auditors

We also surveyed internal auditors at a national Institute of Internal Auditors (IIA)

chapter committee meeting to understand their perceptions of the need for an RPA

framework. We received 56 responses with full or partial data.

Respondents came from

publicly traded companies (22 responses), private companies (22 responses), government or

not-for-profit institutions (6 responses), and self-reported “other” types of organizations (6

responses). There were 37 respondents who indicated their organization had more than 5,000

employees, 11 had between 1,000 and 4,999 employees, 7 had between 50 and 999

employees, and 1 had less than 50 employees. Participant job titles included head of internal

audit, chief audit executive, audit managers, and senior auditors.

Responses to key questions are listed in Table 1. We asked participants if their

internal audit functions had a “written framework that helps auditors decide which tasks they

should and should not automate.” Of the 56 responses, only 5 indicated they had a framework

(8.9%). When we asked how helpful a framework would be on a 7-point fully labeled scale

ranging from very helpful (1) to very unhelpful (7), the mean response was 2.50, which is



We also provide additional evidence later in the paper of the demand for a framework by surveying external

auditors about their perceived need for a framework.

Ten participants did not answer all the questions (i.e., they provided partial data). We include their available

responses in our analysis. However, excluding their responses leads to inferentially identical results.

One concern is that respondents may not have understood what RPA is and thus could not provide accurate

responses to our survey questions. To ensure that respondents have an understanding of RPA, in all surveys we

define RPA and explain how it is different from other forms of automation. Participants also indicate that, on

average, they are planning to use RPA in the future, suggesting that they have some understanding of the

technology. Of the 56 internal auditors surveyed, only four responded that they did not know if their

organization had an RPA framework.

Electronic copy available at: https://ssrn.com/abstract=3651028



between helpful and somewhat helpful and is statistically different than the midpoint of 4 (p-

value < 0.01). Overall, we interpret this evidence to suggest that (1) very few internal audit

functions have a framework for deciding what tasks to automate and (2) most would prefer

having such a framework.

RPA Vendor Frameworks

We approached the three largest RPA vendors (Automation Anywhere, Blue Prism,

and UiPath) and inquired about frameworks they had developed for internal use or that they

used with clients. Each company put us in contact with an experienced professional at their

organization to answer our questions. The three companies indicated they had various tools to

help customers decide what to automate. However, upon review and after discussion with the

RPA professionals at those organizations, we found that these tools were all designed so that

after the user answered a few basic questions, the decision to automate a task or not was

elevated to an experienced RPA professional to decide about automation potential. That is,

the individuals indicated that their focus on helping customers decide what to automate was

centered on helping large-scale implementation of RPA which, in turn, should be decided by

RPA experts or RPA expert teams. They agreed that they did not provide a framework that

would help individual auditors decide how to automate tasks for audit settings without the

help of an RPA expert.

Taken together, the interviews with external auditors, the survey of practicing internal

auditors, and the discussions with the top three RPA vendors revealed that there is a need for

an RPA framework that helps auditors decide what tasks they should automate with RPA.



We also asked participants if they would prefer a general or audit-specific RPA framework. The results

suggest a slight preference for a general RPA framework. In the context of our study, we interpret this to mean

that a framework should be flexible to not only apply to audit settings, but other settings as well. We take this

into account in the development of our framework.

As discussed in more detail below, we incorporate and expand upon what we learned from the vendor

frameworks when developing our framework for auditors.

Electronic copy available at: https://ssrn.com/abstract=3651028

剩余56页未读，继续阅读

身份认证购VIP最低享 7 折!

30元优惠券

weixin_38652636

粉丝: 6

审计任务自动化：机器人流程自动化评估框架

电机控制领域Harnefors观测器的Matlab仿真模型及其在PMSM无感控制中的高效应用

COMSOL相场法在水力压裂模拟中的应用：从单一裂缝到复杂多簇裂缝的数值实现

管道清污机器人sw16可编辑_三维3D设计图纸_包括零件图_机械3D图可修改打包下载_三维3D设计图纸_包括零件图_机械3D图可修改打包下载.zip

keras-3.3.2.tar.gz

C语言课程设计的一些经典项目以及源码.zip

水果采摘机器人sw22_三维3D设计图纸_包括零件图_机械3D图可修改打包下载_三维3D设计图纸_包括零件图_机械3D图可修改打包下载.zip

爬百度文库ppt(1).py

新能源电机sw22可编辑_三维3D设计图纸_包括零件图_机械3D图可修改打包下载_三维3D设计图纸_包括零件图_机械3D图可修改打包下载.zip

基于CNN-LSTM的锂离子电池健康状态SOH估计方法及其Python实现

三相不平衡电压下T型NPC三电平并网逆变器的控制策略与实现

最新资源