A guide to AI in accounting and finance #2: Prompts and challenges
In the first part of this two-part blog series, we cast a lens on artificial intelligence’s adoption, and subsequent transformation, in the accounting and finance function, specifically examining the benefits and outputs of AI technology. Now, in the second part, we’re focusing on prompt examples to apply and pitfalls to overcome – in order to succeed in your AI strategies, goals and objectives.
Without further ado…
Example prompts and scenarios for AI queries
To collect information from various sources, consolidate it and prepare reports and documents for period end (and routine) activities or in readiness for annual auditing from scratch would take an excessively long time. With AI, these tasks happen in literally a matter of minutes. By way of illustration, here are a few real-life scenarios of AI prompts in an accounting and finance setting:
Please summarise the financial position of the firm for January, February and March, covering the specific headings working capital (which should include aged debtors, aged disbursements and aged WIP), client liabilities (which should include client account balances), cash flow movement, fee performance by staff member and profit.
From the invoices report, please identify any invoice which is outstanding on 31st March with an invoice date of more than 30 days. Using the address information in the client details report, please cross reference these client names and provide a list of addresses to send a credit control letter, using the address fields of address 1, address 2, town/city, county and postcode.
Please create a management report which includes the key financials from each month in the quarter ending 31st March, with an executive summary of the firm’s financial position, identifying any concerning large debtor and WIP balances which need to be actioned in the April month end.
Please compare the document titled office receipts report and cross reference with the invoices reports for each month, identifying the average number of days between the invoice date and invoice receipt date. Please then utilise that information to calculate an average number of debtor days on invoices where receipts have been paid against invoices with the same invoice number.
Please review the fee earner summary reports for the three-month period from January to March and identify which staff members have the highest fees billed amount compared with their fees entered budget figure for the same period.
Our firm’s average debtor days are X. How does this compare with the current average in law firms in the UK?
Our EBITDA is 18%. How does this compare in our sector?
Please create a revised and comprehensive checklist for my legal cashiering team to follow as part of robust month end controls. The checklist should identify the tasks required and the sequence of those tasks, identifying any dependencies and also anything required for compliance with the solicitors’ accounts rules and HMRC compliance.
Of course, this is by no means an exhaustive list. To echo the point above, the possibilities really are endless.
Challenges of using AI in legal accounts
In specialist, business-critical accounting and finance roles, it would be remiss not to point out there are concerns about security and compliance risks associated with the use of AI, some of which are outlined here:
Client money is sacrosanct. Protecting client money is the number one priority of any legal accounting professional. Don’t become complacent about data security as it increases the risk of data breaches which is an infringement of accounting rules and regulations.
Wider impact of AI internally and externally to your organisation. The impact on professional indemnity insurance, staff retention, supervision oversight, billing models, pricing transparency and more are still evolving. These factors are likely to be discussed at length over the months ahead in relation to how financial information is delivered to your law firm’s senior leaders, government departments and regulatory bodies. Watch this space!
Historically, the law sector hasn’t been an early adopter of new innovations. AI poses an interesting new chapter for practices as they progress along their AI adoption journey. Companies must strike the right balance between embracing AI’s transformational opportunities and fighting AI’s privacy threats.
Lack of confidentiality within open web-based AI answer engines. To keep your data safe, only use AI systems which operate in your own protected environment such as Microsoft Copilot or functionality within your case and practice management software, for example LEAP’s Matter AI. That’s because open web models can memorise and reveal personal information by lacking the necessary safeguards to prevent unauthorised data access.
Inability to deliver region-specific results. If you’re in the UK, you don’t want outputs referring to other currencies and VAT/GST rates or other nuanced differences. If your AI system knows your location, results pertain exactly to your region which is much preferable and actionable.
As we've explored in part two, the way you interact with AI is just as important as the technology itself. Thoughtfully crafted prompts unlock impressive efficiencies, while a clear-eyed understanding of potential issues ensures AI’s operated safely, compliantly and effectively – which is paramount on the legal cashiering stage. As AI continues to evolve, law firms embracing it with both curiosity and caution are best positioned to lead the way. At Jayva, we’re here to help you strike that balance and get the most out of the tools at your disposal.
Even more help from Jayva
An expansive range of earlier blogs and guides on the subjects of AI, legal technology and financial management are available at www.jayvaglobal.com/news and www.jayvaglobal.com/articles respectively.
With an equally broad catalogue of consultancy and training services empowering you to embrace technology, invest in your people, improve your data and realign your processes, find out more and get in touch at www.jayvaglobal.com/contact-us. Whether it’s upskilling on AI, mentoring cashiers, unravelling and rescuing accounts, setting up bespoke reports, whatever, we’ve got you covered.