AI still playing ‘backstage role’ in accounting
Accountants are optimistic about AI and yet few are investing in the technology. Among those who are using AI, it is mostly for non-accounting purposes.
Accountants are more likely to not use AI at all than to use it for accounting-specific functions. Until firms increase their spending on AI training, this is unlikely to change. These and other findings were shared in Karbon’s The State of AI in Accounting Report 2024.
The report, which gathered global survey data in December of last year, found that 22 per cent of accountants were not using AI in their work. Of those who were making use of the technology, it was typically for non-accounting functions.
Fifty-nine per cent of accountants reported using AI to craft emails and to inform their writing tone, making it the most common deployment for accountants followed by task automation (36 per cent) and research (31 per cent).
Comparatively few were using it for accounting-specific functions including financial forecasting and analysis (11 per cent), client support (10 per cent), and tax preparation and audits (6 per cent).
“AI needs to be incorporated into actual tax and accounting software to streamline real production work, otherwise it’s mostly for marketing and internal document generation,” one respondent told Karbon.
Investment in AI training appears to scale with the size of an accounting firm but only up to a point, after which the trend reverses. Firms with more than 200 employees were less than half as likely to offer AI training than firms with 51–200 employees. This trend has been dubbed the ‘AI paradox' in accounting.
The report suggested that the drop off in training is symptomatic of a slower pace of change within firms of 200 staff or more.
“Beyond a certain size, firms encounter new complexities – hierarchical structures and strategic inertia that may impede the adoption of nimble, innovative practices like AI training,” said Karbon.
The trend is surprising given the wealth of research suggesting early adopters of AI were disproportionately larger. More recent research, however, such as the work conducted by Piper Frangos of the Hult International Business School, has established that internal organisational structures and leadership play a determinative role in AI readiness.
A literature review conducted by Dr Frangos found that agile leadership, for example, has an outsized role to play in corporate AI adoption and proactive change management.
Despite the growing awareness of the benefits of AI in accounting – for instance, 54 per cent of accountants said that firms that do not use AI are less valuable – trepidation continues to hamstring its adoption, said Karbon.
Seventy-six per cent of respondents, for example, said they were concerned about data security when evaluating AI tools. Given that accounting is among the most vulnerable sectors to cyber attacks the trepidation is not without merit.
“I believe AI is a massive threat to the workplace globally and I believe the possibility of cyber intruders will become even more threatening,” said one respondent.
A further 56 per cent of respondents said they were concerned about losses to human touch in the delivery of accounting services and perceived costs to human relationships.
Undergirding these concerns is a sense of disconnect from the nature of AI developments as 46 per cent of respondents found it a challenge to keep up with AI technology developments.
Nearly 90 per cent of those who reported being ‘scared’ of AI reported having no AI training. Those who reported feeling ‘excited’ about the technology were more likely to belong to firms that were developing their AI training or already offering it.
“The best place to begin closing the current training and knowledge gap is with exploration. Team members should be encouraged to test, experiment, and learn,” concluded Karbon.
