ICAEW sees the opportunity that generative artificial intelligence (GenAI) technology represents for finance practitioners to accelerate routine processes, enabling attention to be focused on interpretation, decision-making and adding value. According to Peter Beard, ACA, Director of GenFinance.AI “GenAI is transforming our work by converting many manual tasks into more efficient, semi-automated workflows, enabling much faster preparation and review.”

How can you use GenAI in your workflows?

“Period-end close has long been associated with late nights, repetitive reconciliations and a heavy reliance on spreadsheets,” says Beard. “GenAI promises to ease the burden at four key stages of the process as we outline in our GenAI Accelerator Programme.” 

The second course in the programme launched at the end of September and focuses on financial operations and decision making. It includes a whole module on financial accounting and period-end close. offering step-by-step guidance and real-world examples on how to integrate GenAI into period-end close processes.

It focuses on how GenAI can be used at four critical points in the process:

1. Initiation

This is where an economic event occurs, such as a sale, purchase or expense. This stage is about identifying that a transaction has happened which requires accounting recognition.

Traditionally accountants would spend a long-time reading emails, PDFs and other documents to identify and code transactions. However, GenAI can scan documents, summarise emails and suggest classifications based on the content it has been provided. 

GenAI can help by identifying and classifying events from unstructured data. For example, it could review emails to check for any new transactions or an internal communication about a new purchase. A quick sense check is then needed to assess whether this event requires accounting recognition. 

2. Recording

This step is about capturing the transaction in the books using double-entry bookkeeping, creating a journal entry that reflects the debits and credits to the appropriate accounts.

Traditionally for new transactions, accountants would spend time interpreting contracts or agreements to determine the correct debits and credits. However, accountant can now simply describe the transaction in plain English to a GenAI model which can then draft the correct multi-line journal entry based on accounting standards.

For post-journal entries practitioners would manually enter information into general and subsidiary ledgers, which was highly prone to data entry errors. However, GenAI can suggest postings from a list of transactions, ensuring consistency and accuracy.

GenAI makes this step faster by translating a natural language event into a structured journal entry. For instance, if it is provided with a company’s chart of accounts it can suggest the correct nominal codes for a transaction, turning a manual data entry task into a review and approval process.

3. Processing

Processing involves the critical work of verifying accuracy through reconciliations and making adjustments, such as accruals, prepayments and depreciation.

Traditionally to reconcile bank accounts, accountants would print statements and tick off each item against the cash book - a tedious, line-by-line process. However, with GenAI practitioners can upload both datasets and the model can identify matched items, discrepancies and unrecorded transactions.

In the past, to reconcile control accounts accountants would reconcile VAT, payroll and suspense accounts often using complex VLOOKUPs in spreadsheets. However, when prompted GenAI can analyse control account movements, match sub-ledger data and drafts reconciliation summaries, and flag unusual items. 

Previously when investigating ledger inconsistencies, practitioners would have conducted a ‘needle-in-a-haystack’ search through transaction listings to find the source of an imbalance. But GenAI can analyse large datasets to quickly pinpoint transactions or patterns that are causing the discrepancy, saving large amounts of time.

GenAI is able to perform high-volume reconciliations in minutes, highlight discrepancies and even calculate depreciation when provided with the policy and data. This means, instead of wrestling with spreadsheets, accountants can focus on investigating issues and applying judgement.

4. Reporting

The final stage of the period-end close process, focuses on producing financial statements and drafting commentary to explain results to stakeholders.

Traditionally when preparing a trial balance accountancy teams have manually summed debit and credit columns to ensure they balance and conduct a search for the error if it didn’t. Now the trial balance can be uploaded to a GenAI tool which will verify if it balances and help identify accounts with anomalous balances. 

To perform an analytical review, accountants calculate key ratios and compare current period figures to budget/prior periods in Excel. GenAI tools can perform variance analysis on a trial balance and draft initial commentary on key movements. 

Finally, draft management reports are often created based on data by cost centre or department, involving extensive spreadsheet work and narrative writing. However, GenAI can summarise data by specified dimensions and draft narrative reports on departmental performance against budget, saving vast amounts of time.

It can then assemble first drafts of reports, calculate ratios and perform analytical reviews. The tools can also draft footnote disclosures, based on provided data and the relevant accounting standard. 

By automating initial commentary on budget variances, accountants can concentrate their efforts on interpretation, governance and providing the strategic narrative.

Seeing the benefits in practice

A financial accountant could spend hours at month-end reviewing lengthy supplier contracts and complex spreadsheets to calculate the accrual for services received but not yet invoiced. The process is prone to interpretation errors and is a significant bottleneck.

However, when supported by GenAI, they can upload an anonymised service contract with the instruction: “Based on this contract and our policy of accruing over the service period, calculate the correct accrual for the month of June 2025.” 

The GenAI tool should then extract the relevant terms and provides a calculated figure with a clear explanation. 

“The period-end close will always require professional oversight, but GenAI changes the nature of the work,” says Beard. “This technology can help accountants save time on the manual laborious tasks in the preparation work, so that they can concentrate on interpreting the data, investigating issues and applying governance and judgement.”