Accountants and finance leaders adopting AI are beginning the next chapter in the profession’s tech uptake. It’s a tale that is anything but straightforward.
Although counting devices such as the abacus have been in place for thousands of years, the arrival of Italian mathematician and Franciscan friar Luca Pacioli and his work on double-entry bookkeeping in 1494 is when our story really starts.
The Industrial Revolution resulted in more complex accounting but mechanising the process was only realised in the Victorian age when James Ritty’s cash register was launched in 1879. It was followed by William Burroughs’s adding machine – which could have been scripted by HG Wells – the following decade.
‘Once we let finance people try it, we discovered there was demand’
Punch-card machines moved things forward, and by 1928 an IBM tabulator could handle 100 cards a minute. But it was the advent of mainframe computers in the mid-20th century that brought accountancy technology into the modern age.
Great tech breakthrough
Bob Frankston and Dan Bricklin delivered the great tech breakthrough for accountants in the winter of 1978-79 with the creation of VisiCalc – short for ‘visible calculator’ – the first spreadsheet computer program for personal computers. Their vision turned the microcomputer from a hobby for computer enthusiasts into a serious business tool for wide use.
The success of their company, Software Arts, was not a given, as the Apple II that VisiCalc was designed to run on had a tiny memory. ‘When we first started work on the project we had modest expectations, but once we let finance people try it, we discovered there was demand,’ Frankston says, conceding that computer programmers were ‘underwhelmed.’
But succeed they did. Since 2010, 17 October – the date VisiCalc was launched in 1979 – has been celebrated as Spreadsheet Day, if you weren’t already aware.
Easy as 1-2-3
The fortunes of VisiCalc would depend on how soon the formula could be improved on. ‘It was surprising how long it would take for people to copy the idea,’ says Frankston.
All of three years, to be precise. That’s when Mitchell Kapor, head of development at VisiCalc’s distributor VisiCorp, walked out the door and with Jonathan Sachs launched Lotus 1-2-3, an electronic spreadsheet designed to run on the new IBM PC.
Lotus raked in sales of US$53m in its first year, making it larger than Microsoft
Kapor’s formula for making software as powerful as possible yet easy to use made it the bestselling personal computer software of its day. ‘The power of the product owed to its versatility: the ability to create spreadsheets, graphs and do data analysis in the same program,’ he says.
‘But it also represented a huge leap forward because of its greater speed, the ability to develop much larger models by taking full advantage of the IBM PC’s memory capacity and the ability to automate repetitive processes through the macro language. All this was done while also enhancing ease of use through the menu, prompting and help systems.’
So successful was the program that Lotus raked in sales of US$53m in its first year, making it larger than Microsoft at the time.
Lotus 1-2-3 was disrupted by the introduction of Microsoft’s Windows operating system and its suite of office-productivity software, including Excel, which remains the world’s most popular spreadsheet program. That’s despite a resurgence in the 1990s with the launch of software platform Notes, resulting in Lotus being snaffled up by IBM in 1995 for US$3.5bn.
Age of AI
The need for more comprehensive financial management tools by the 1990s resulted in the growth of integrated accounting software by firms such as Sage and QuickBooks that combined ledger, payroll and inventory management functions.
The new century ushered in the cloud computing revolution, with providers such as Xero and QuickBooks Online enabling users to access their financial data anywhere with an internet connection, making financial data more accessible.
Most recently, the arrival of AI is set to turbocharge accounting software by automating daily tasks such as data entry and reconciliation, reducing the likelihood of human error. At the same time, machine-learning algorithms facilitating predictive analytics can deliver insights into future financial trends.
Slow start
But things could have been different decades ago, says Professor Miklos Vasarhelyi, KPMG distinguished professor of accounting information systems at the Rutgers Business School in the US, who published six books on AI in accounting between 1989 and 2004. ‘We stopped the series as no one seemed interested except academics,’ says Vasarhelyi, who also pioneered continuous audit reporting, a concept that has started to gain traction again in the last two or three years.
Alan Sangster, professor of accounting history at the University of Aberdeen, says that accountancy’s failure to adopt AI technology back then might have been due in part to the profession’s concerns its status might be undermined.
‘There was even a hesitation by the UK government about using computers for tax returns’
‘I think that technology was always viewed as a threat. In the late 1980s I wrote how expert systems were going to take away the skill set of accountants and accountants were going to be replaced by machines,’ he says.
‘This professional bypass was potentially on the horizon; now it is very much present. There was even a hesitation by the UK government in the 1990s about using computers for tax returns.
‘Expert systems began to make an impact, and then they disappeared. They just went off the agenda,’ Sangster continues, adding that blockchain technology is poised to revolutionise accounting practices once again by ensuring financial transaction integrity, through reducing the risk of fraud.
But he insists that an AI-led approach will make accountants feel threatened: ‘It’s not a quick fix; it will take an awful lot of fine tuning.’
Too much tech?
Karen McBride, professor of accounting at the University of Portsmouth, believes that the evolution of tech has brought advantages in increased efficiency, potential for real-time reporting and better data analysis capabilities. ‘Improvements in automation and AI have replaced basic accounting tasks; for example, bookkeeping and reconciliations are used in audit testing, leaving accountants to focus on advisory services and adding value,’ she says.
‘There is a risk of an acceptance of a given answer without a critical, questioning approach’
She adds that junior staff who might have developed skills carrying out these routine tasks no longer experience often repetitive tasks, which can enhance their training experience, but she warns this may also leave them lacking in practical experience.
‘A development of professional judgment based on learning underlying accounting practices and principles could give way to a reliance on tech-generated solutions. As with GenAI, there is a risk of an acceptance of a given answer without a critical, questioning approach,’ McBride says.
The fast pace of technological advancement may also exacerbate inequalities within the profession, asserts McBride. ‘Larger firms have more resources to invest, leaving smaller firms behind and potentially increasing a digital divide. There may be challenges in reacting quickly and thoroughly to ensuring data privacy and security,’ she says.
‘I did not see a backlash, but businesses and governments are not interested in an anachronistic set of rules that does not represent their operations very well,’ says Vasarhelyi.
‘For running their entities, they need much more modernity in their reporting, a change towards a modern entity measurement and close to real-time validation. We need continuous AI-based reporting, monitoring and assurance.’
