Introduction
Accountants should use Generative AI (“GenAI”) at their work. Current GenAI is particularly efficient with regard to language / text tasks which constitute a significant proportion of each accountant’s work.
Key tasks GenAI is very good at can be summarized as follows:
- “Summarizing (e.g., summarizing user reviews for brevity)
- Inferring (e.g., sentiment classification, topic extraction)
- Transforming text (e.g., translation, spelling & grammar correction)
- Expanding (e.g., automatically writing emails)” (source: deeplearning.ai)
For brief 1-hour intro into effective use of GenAI I would recommend a course from deeplearning.ai.
OpenAI is widely known and has entered into significant partnerships, so from my point of view it will probably cement its position as the main GenAI provider. However, as of the date of writing, I am currently using perplexity.ai which is more reliable than OpenAI based on my user experience. Whichever GenAI tool you decide to use, pin it to your web browser so it always opens automatically. Below are very practical tips as to how I have been using the tool for my day-to-day tasks.
Accounting Memos, Manuals, Training materials
Accountants can leverage GenAI to streamline the creation of accounting memos, manuals, and training materials. For example for accounting memos, if accountant provides the basic description and purpose, GenAI can:
- Analyze the situation: GenAI can review an input from the accountant regarding the topic and identify potential revenue recognition issues for long-term contracts
- Research applicable guidance: It can search and summarize relevant IFRS 15 guidelines for revenue recognition
- Draft the accounting memo: GenAI can create an initial memo outlining the proposed treatment for recognizing revenue over time for a construction project
- Suggest appropriate accounting treatments: It can recommend journal entries for recognizing revenue, associated costs, and relevant balance sheet positions as performance obligations are satisfied
This approach significantly reduces the time spent on initial drafts, allowing accountants to focus on review and refinement.
Internal Controls Procedures & Documentation
GenAI can be used for analysis and documentation of internal controls:
- The accountant inputs key process descriptions, for example, for Purchases & Payables cycle, description of the process from creating an order through receiving an invoice to payment approval and disbursement
- The accountant then request GenAI to list up the key risks (LSPMs), e.g., for Accounts Payable. GenAI might identify risks like unauthorized orders, duplicate payments, and similar.
- For each risk, the accountant may request GenAI to generate up to 5 controls, including their:
- Control titles, for example “Three-way match verification”
- Detailed descriptions, for example “Compare purchase order, receiving report, and vendor invoice before payment approval”
- Responsibility assignments, for example, Accounts Payable Clerk
- Control frequency, for example, multiple times a day
- other attributes, as relevant
After this, the accountant can review the output and tailor it.
Presentations and Reports
Accountants often prepare various presentations and reports – audit reports, management reports, financial statement footnotes, investor relation materials, internal audit reports, due diligence reports, and similar. These often require one of the core activities described above – summarizing and expanding – in which GenAI is excellent, so use of GenAI saves a lot of time.
- The accountant can request condensing lengthy descriptions into concise bullet points for PowerPoint slides, for example, summarizing a 500-word explanation of the current quarter EBITDA development into 3-5 bullet points for an executive presentation
- The accountant can instruct GenAI to draft initial versions of management reports and audit committee presentations, for example, creating an initial draft of a quarterly financial performance report, including key metrics like gross margin and operating expenses
- Request summarizing key internal controls findings and recommendations for client reports, such as condensing a lengthy audit documentation into a 1-page executive summary highlighting material weaknesses in internal controls, instead of coming up with wording themselves
- Instructing GenAI to provide an initial disclosure drafts for financial statements footnotes compliant with specific accounting standards (e.g., IAS 19), and giving it the basic information related to the disclosure, such as discount rates, etc., possibly also providing GenAI with the prior period disclosures as a reference point so that output is more relevant
Other
Beyond these specific tasks, GenAI offers versatile support for accountants in various daily tasks:
- Drafting professional emails: Request GenAI to compose an initial email draft to a colleague or a client explaining the impact of an accounting standards/interpretation on their new transaction in the financial statements
- Suggesting research starting points: Request GenAI to provide a list of relevant IFRS standards and interpretations for researching the accounting treatment of cryptocurrency holdings
- Providing initial consultations: Offer a preliminary analysis of whether a series of transactions qualifies for hedge accounting under IFRS 9. However, output has to be critically reviewed as oftentimes such consultation are incorrect, based on my experience. Yet, they still provide useful ideas
- Reviewing work: Request from GenAI a critique on a drafted uploaded memo on the accounting treatment for a complex business acquisition, and request it to highlight areas that need more support or clarification, and check the grammar errors
Additional areas – near future
These are further use cases that are currently evolving and it is expected that they may be more reliable in the near future, but based on my experience as of today they still require a significant work from the accountant, so they are not saving so much time yet.
- Financial Statement Analysis: GenAI will analyze financial statements, identify trends, and generate insights for management discussion and analysis sections
- Audit Planning: Accountants may use GenAI to assist in risk assessment and audit planning by analyzing historical data and industry trends
- Tax Planning: GenAI can help identify potential tax-saving opportunities by analyzing a company’s financial data and current tax laws
- Fraud Detection: By inputting financial data and transaction patterns, GenAI can help identify potential red flags for fraud. The accounant could upload a whole general ledger with the booked transactions, and request GenAI to flag suspicious items
- Budget Forecasting: Accountants can use GenAI to generate initial budget forecasts based on historical data and current market conditions
- Regulatory Compliance Checks: GenAI can assist in ensuring compliance with various accounting standards and regulations by reviewing financial statements and disclosures
- Continuous Learning: GenAI can provide accountants with summaries of new accounting standards or tax laws, helping them stay up-to-date with industry changes
- Data Visualization: Accountants can use GenAI to suggest appropriate data visualization techniques for financial reports, making complex information more accessible. However, numerical skills are not yet advanced, although recently there seems to be some progress in this field too
- Process Automation: For example, if public GenAI tools are more reliable, it can be foreseen that small business will directly connect invoice processing to ChatGPT – instead of software providers – and ChatGPT will read all incoming invoices, and automatically scrape the key data into an Excel file which will be ready for an input into bookkeeping software
Privacy and confidentiality of data
Accountants need to be mindful about the privacy and confidentiality of the data, so private GenAI tools are the right choice – such as those where your organization has a partnership with the GenAI provider, and ensures that the data is securely stored and not shared outside the organization. Alternatively, if a public GenAI is used, the accountant needs to be very mindful and anonymize any sensitive data that should not be leaked, before using the tool – particularly those that relate to names, dates, locations, sensitive descriptions, and the amounts.
Conclusion
While GenAI’s output may sometimes be incorrect, it serves as a valuable tool for generating ideas and improving efficiency for accountants who review the output and improve it as needed. It is already essential that the accountants use it at their work daily because GenAI significantly cuts down time spent on activities that are related to language and text in general.