Nothing to say, yet
Transcription
Artificial intelligence (AI) automation in accounting is revolutionizing the industry. AI utilizes various technologies like data mining, machine learning, and speech recognition to efficiently analyze large data sets. It can perform almost half of accounting tasks, saving trillions of dollars annually. AI is also valuable for risk assessment and has been implemented by major accounting firms. However, there are weaknesses, such as the potential for tampering and cyber attacks. Auditors can mitigate these risks by ensuring quality data input and implementing controls. Training AI to detect malware and following governance frameworks can enhance security. Increased regulation and guidance are necessary for safe and effective use of AI in accounting. Combining frameworks and AI-powered systems can empower accountants to monitor and analyze business operations more effectively. Today, we'll be talking about AI automation in accounting. Artificial intelligence is the ability of computers to solve problems and complete tasks using logical thinking and rational actions like humans do. Artificial intelligence is a culmination of connected technologies to include data mining, machine learning, speech and image recognition, and semantic analysis. Artificial intelligence is a new technology that provides many innovative opportunities. However, as with all new technologies, challenges have presented themselves, and internal controls are needed to maintain the technology's integrity, transparency, objectivity, and overall effectiveness. The collection of such large data sets has made AI applications incredibly valuable to businesses who are looking to efficiently analyze all of that data. Types of AI, such as machine learning, can assist the business in training a software model to make reasonable predictions from the collected data. AI applications, like robotic process automation, minimize or eliminate the need for humans to perform lower-level accounting tasks, like preparing tax returns and financial statements. AI is actually capable of completing almost half of accounting tasks, which can save trillions of dollars in the global workforce annually. In fact, because of AI's superior ability to take advantage of algorithms to identify patterns and irregularities in entire sets of transactions, it makes for a great risk assessment tool for auditors. The four largest international accounting firms have already implemented AI into their audit and accounting processes. Deloitte's Argus AI application uses cognitive thinking to learn from each employee who inputs information to the application and takes advantage of natural language processing to identify and extract important accounting information from any document that has been uploaded. Like Deloitte, KNPG uses AI to enhance their decision-making process and business operations by using natural language processing to extract information from sources like customer calls, which is then used to generate predictions of future trends and business events. PwC uses their Halo application to integrate all transactions, which increases the quality of testing and risk assessment. EY uses deep learning artificial intelligence to examine unstructured data, such as emails and audio files from conference calls, which reduces the time needed to review audit documentation so that auditors can focus on higher-level analysis. Overall, AI has essentially made it possible to engage in a continuous audit process instead of an annual or periodic audit process. Even with the advantages that AI brings to the table, there are still weaknesses within this new technology. AI systems can be tampered with by malicious users to hide financial deficiencies during an audit. Hackers are also able to inconspicuously infiltrate AI systems during intelligent cyber attacks. During these attacks, they can use AI-learned behavior to stay there for extended periods of time and perform tasks like setting up back doors to aid in future attacks. One of the most vulnerable AI components is image-based training sources because hackers can use the fast gradient sign method to compromise source images and corrupt system algorithms. All of these weaknesses stem from the fact that artificial intelligence is only as reliable and accurate as the data that is fed into it. Auditors can lessen these risks by paying special attention to things like supporting documentation and user interface logs to ensure quality data is being put into the system. Data entry input controls such as range checks, validity checks, completeness checks, reasonableness checks, and closed-loop verifications can ensure the system is properly vetting data and the related source documentation. In addition to that, the implementation of an access matrix acts as a segregation of duties for system users to prevent unauthorized access and possible manipulation of data that can skew reporting results. AI can also be trained to detect malware by running samples through the system using deep learning. In the event of an intelligent cyber attack, corrective controls like recovering correct data with backup files help to quickly get the system secure again. Increasing the regulation and guidance on best practices will be necessary to ensure safe and effective use of AI technology in the future. In the meantime, accounting professionals can use existing governance frameworks like the Control Objectives for Information and Related Technology and the International Organization for Standardization 27000 series to implement best practices and controls for IT security and management. Combining COVID and ISO frameworks, AI-powered IT security systems, proper controls, and cross-departmental support can empower more accountants to leverage the benefits that artificial intelligence has to offer so that they can elevate their ability to monitor and analyze business operations on a higher level.
