At a glance
- Industry leaders advised accountants to disclose AI usage to clients.
- Automation is reshaping accounting practices.
- Firms encouraged to integrate AI with internal processes first and develop clear policies for organisational use.
The need for transparency should not discourage accountants from embracing the technology given the productivity benefits, industry leaders told the Accountants’ Guide to AI & Digital Assets online conference on 15 November.
Accountants grappling with disclosure requirements for outsourcing and digital services were encouraged to update their service agreements and privacy policies to reflect AI use. “Disclose, disclose, disclose, and understand how AI will exist in your organisation,” said Jade Peace, Chief Operating Officer of Lextech.
“If the service that you’re providing is financial advice and you are using an AI tool to deliver that financial advice, you must disclose that to the client at any given time, regardless of where you are.”
KPMG has waded into the AI space, developing a tax research tool that was “designed by a tax professional, for a tax professional, that only a tax professional could use”, according to Ed Hewings, the firm’s Director of Artificial Intelligence & Digital Solutions.
“Every piece of AI that’s out there at the moment is the worst version it’s ever going to be,” he said. “It’s only going to improve from this point onwards.”
AI applications transforming accounting operations
The conference session, moderated by Lisa Greig FIPA FFA from Perigee Advisers, explored how practical applications are changing the way accounting practices operate.
Adrian Smith, General Manager at Run Gopher, used AI to keep projects on track and maintain visibility. He described how AI meeting tools now generate minutes, create action items and suggest agenda points based on previous discussions.
“Just try something,” he said. “Test it out, see if it works. If it doesn’t, try the next one. It keeps you on track.”
While attendees were encouraged to dip their toes in the AI waters, an over-reliance on the technology also had drawbacks. Peace cautioned “it isn’t going to be the panacea”.
For instance, the presenters cited legal cases in the United States that had collapsed due to excessive reliance on AI research. Automated systems that amplify rather than correct poor data have caused headaches for companies.
1. Start slow with AI adoption
With AI taking over basic tasks, the profession is exploring the implications for teaching aspiring accountants foundational skills.
Universities are already adapting, according to Peace, who sits on several Australian university boards. AI now features in accounting curricula, reflecting the growing importance of technological literacy together with financial expertise.
The speakers recommended companies start integrating AI in their internal processes where mistakes would not impact clients. For many, meeting documentation is an easy entry point.
Smith described how AI tools transcribe discussions and generate action items automatically. “As soon as the meeting finishes, they get the same notes that we do with a transcript of the meeting, all the bullet points,” he said.
The technology tracks commitments outlined in meetings. ‘It creates action items based on who has said what. That’s all very transparent. Means that we can go back to notes.”
2. Human element critical
Human oversight helps with accuracy of information, data and processes. Peace advised firms to implement AI policies that clearly define the parameters of the technology’s use. “For example, in scope is the use cases ABCD, out of scope is the use cases D, E, F, whatever you determine is out of scope or is not acceptable,” she said.
This guidance should be communicated to all employees. “Be really open and transparent to your employees, especially,” Peace said. “Communicate that clearly to the rest of your team members, because ultimately, in this age, it will prevent you from having a notifiable data breach.”
3. Treat AI like a new employee
Smith offered practical advice for working with AI systems. “Pretend that you’ve got a new employee sitting next to you,” he said. “The reason we’re doing what I want from you is this. Ask me any questions.”
This allowed for further refinement of AI interactions to produce the desired outcome. “You can go back to AI and say, repeat back to me what I’ve asked you,” he said. “It will say, I’m doing this. This is the outcome that we’re wanting to achieve.”
AI tools are developing in sophistication at a time of regulatory complexity. Finding the balance between technological advancement and professional judgment sets companies up for success.
“Have fun with it,” Smith said.
More information on the IPA’s 2024 National Congress on 27-29 November available HERE.
