Skip to content

Artificial Intelligence in Auditing

AI- A groundbreaking revolution

From spell-check to autonomous cars to facial recognition, Artificial Intelligence (AI) is something that smoothly dwelled into our life affecting our regular patterns and habits.By amplifying our abilities and providing a simpler user interface, AI is likely to become the most important human collaboration tool ever created.

There are various notable breakthroughs in AI in the recent times such virtual assistants like Apple’s Siri, Google Assistant and Amazon’s Alexa that have redefined human-data interaction and automation of tasks; Hanson Robotics had created the very famous Sophia, a learning robot who received citizenship in Saudi Arabia; With IBM Watson for Health, clinicians can be provided with evidence based treatment options, patient engagement capabilities and novel drug targets; Kensho which is an intelligent computer system has been widely used by stock traders to automatically analyze portfolio performance and predict market changes and many more.

These exponential technologies are helping us speed towards a world of abundance. With AI in its infancy stage, whatcould we further automate in the future? How can technology be used in the wide area of Auditing and Taxation? And what effect would it have on the human experience?

The inclusion of artificial intelligence in the field of auditing has predominantly increased over the time. AI can reduce the amount of noise that auditors have to wade through before they can assess and draw imperative conclusions.Algorithms can be created to evaluate entire population instead of a sample, eventually increasingthe effectiveness of audit.These algorithms help in analyzing every transaction and identifying areas requiring additional attention, thereby helping auditorsto focus efforts where the risks lie.

 

AI entering the world of Accounting, Auditing, Advisory and Tax:

Accounting : The global giant EY (Ernst & Young) has recently applied AI to the lease accounting process.AI is used to streamline data capturing from numerous contracts by identifying relevant clauses for accounting such as dates, payment amounts, renewal etc.

In future, we may witness AI in accounting which has the potential to pass entries andcalculates inventory values/ depreciation schedules based on the inputs given, which in this case can be the purchase invoices.Accountants may only need to be involved intraining or testing models orcontrolling the inputs oroutputs, such as exception-handling etc. If AI is actually deployed in the accounting field, it has the potential to be recognized as a huge breakthrough.

 Statutory Auditing:

For example, a laboratory experiment (Swinney 1999) helped in the design of a fuzzy logic expert system based on decision rules for the evaluation of the entity's going-concern status and for assessment of materiality.

With the help of deep learning, numerous audit tasks such as reviewing source documents (e.g., bank check, deposit slip, sales invoice), processing paper work, analyzing emails, press release, news, and extract metadata about the clientand analyzing financial statements can be automated. This indirectly, reduces the turnaround time for various audits. Deep learning has the potential to identify the smallest of intricacies, which may be overlooked by an auditor.

Revolutionary drone technology in the Stock count proceduredepicts the capability of AI intransforming areas of manufacturing, supply chain and inventory management. These drones are embedded with sophisticated AI systems to suit the needs of the companies. This was, the cumbersome physical process of the stock count, can be performed from any nook and corner. Not only has this simplified the entire process, it also has reduced the scope of any clerical errors, generally arising from the manual process.

To improve audit quality, EY devised an accounting fraud prediction model. By applying the module to a company’s previous financial statements and combining similar industry insights, it can determine if the FS could contain a misstatement. And this model gets smarter ad smarter with each piece of input given to it.

KPMG, in March 2016, announced its combination with IBM Watson to use cognitive computing in its professional offerings. For example, through application of such technology across a bank's commercial mortgage loan portfolio, more detailed and comprehensive understanding could be obtained on the bank's credit files and potential audit exceptions based on loan grading.

Advisory: During the process of Mergers & Acquisitions, the assessment of potential procurement synergies requires looking at millions of unorganized accounts payable and receivable data. This humungous process could become a nightmare if done using spreadsheets and pivot tables. In the age, AI in its minimal stage is being used in the form intelligent classification engines, which compile the meaningful information swiftly and accurately.

At Deloitte, the work that once took four to five months to complete can now be done in a week, indicating the possibility of analysis during the evaluation stage of the deal.

We may witness that AI in future could even analyze what could happen to a business’s tax risks and overall structure based on the coming election or other such economic factors. Imagine instead of legal team spending hours on contracts, if an AI tool could be able to quickly identify thorny legal problems or taxation risks.

Internal Audit

For instance, in credit scoring, the customer with a long history of maintaining borrowings without defaults is generally classified as “low risk.” However, what may be unseen is that, such mortgages have been supported by substantial tax benefits that are nearing expiry. Predictive machine-learning models with access to the right data can find the hidden patterns in the dataset.

AIis,is an AI software, used by EY to identify client’s fraudulent invoicesfrom millions of invoices. The value addition being a help in avoiding the serious consequences that result from violating sanctions, anti-bribery regulations, or other aspects.

In the possible next digital frontier — artificial intelligence, internal audit cannot be left behind. Internal audit is expected to provide assurance over AI risk management, governance, and controls.

Tax: Natural language generation (NLG) is used by Deloitte for the creation of text by computers, in its tax business. It processes numeroustax returns annually for clients’ employees’ or other complicated financial situations.

Using NLG, Deloitte creates detailed narrative reports of individual tax returns. During consultations, thetax professionals rely on such reports forbetter financial advice to clients.

We might see intelligent algorithms in future, which may analyze employee expenses and determine the availability oftax deductions.

 

AI: Boon or a bane?

One of the major reason for discouragement of AI is the threat it poses in terms to replacement of work done by Chartered Accountants.

As per SA 200, the auditor is told to exercise professional judgment and maintain professional skepticism throughout the planning and performance of the audit. Professional judgmentand skepticism are cultivated from practice and is something that isn’t taught. If we think about it, AI infact helps the auditor to exercise these two with at most efficiency.

Our intuitive thinking is exceptionally powerful in terms ofdepicting quick learning and high levels of flexibility.

Professional are not just expected to know the rule book to taxes and accountancy, but are expected  to read between the lines, follow the intention of law in case of complications, identify the ways of tax planning .The technology cannot replace the need for expert knowledge and decision-making.

Other disadvantages are in terms of difficulty in governance of AI, requirement of high quality of data maintenance and organization and data security.

Howeverthe benefits of AI outweigh such difficulties. The benefits include automation of routine decisions thereby saving time of the organization,Improved forecasting and analysis for better consultation, adaptive learning and many more.

To conclude, reluctance in acceptance and usage of Artificial Intelligence in the profession would definitely not prove to be a preferred course of action and in turn a mammoth loss.

In the coming years, we could be witnessing the recruit of ‘data assurance associates’ or ‘blockchain auditors’, instead of mere audit graduates of standard curriculum.

A challenge will always remain for the profession to keep up with the developments. Let us all be ready to optimize what the future in AI has to offer and keep in mind- “You need to innovate, or you will become obsolete”!

 

Author

Meghna

Meghna is an Audit Analyst at KVA. She has worked on varied assignments covering risk management, statutory compliance and taxation advisory. The evolution in the current e-generation fascinates her and she is eager to explore the advancements in such field.

Leave a Reply

Your email address will not be published. Required fields are marked *

Show Buttons
Hide Buttons