Generative AI through the lens of accountants and tax practitioners
Key Takeaways
- Since its launch in November 2022, the freemium version of OpenAI's generative AI tool ChatGPT registered over 100 million users within 3 months. It has been estimated that, by March 2023, 14% of American adults had tried ChatGPT.
- Accountants and tax practitioners in Australia who are considering using generative AI must navigate the legislative provisions of the Tax Agent Services Act 2009 and the regulatory guidance from APES 110 Code of Ethics. Updates to engagement letters may be necessary.
- Possible use cases for AI tools for accountants and tax practitioners include streamlining research on tax codes and accounting standards, identifying tax advisory opportunities, strengthening customer relationships and optimising firm management and operations. However, not every generative AI tool is the same and an accountant or tax practitioner should only use an AI tool where the source material is known and validated.
- There are perceived problems with the method around how open source AI tools handles calculations. Accountants and tax practitioners need to keep this in mind as it is a current core strength for professionals. A series of case studies uncovers and reviews these problems.
- Case studies analysis of sophisticated and complex tax problems reveals that open source generative AI models are not up to standard due to technical errors, no critical filter being applied and a general sense of simulating a response rather than understanding the response.
- The continuing development of soft skills is still required for accountants and tax practitioners to be regarded as the trusted advisor for their clients or employers. Taking a client-centric approach will ensure your services are retained throughout the business lifecycle.
- While AI isn't replacing accountants and tax practitioners, it can act as a powerful companion, helping guide decision-making, augment existing capabilities, and highlight valuable data that previously may have been missed. With AI and automation working together, professionals can focus their professional judgment and expertise more directly on complex situations, enhancing their value.
eBook - Table of Contents
- Executive Summary
- Introduction
- Glossary
- What is generative AI? How has its current iteration been received?
- How does generative AI work within the Professional Code of Conduct and the Tax Agent Services Act 2009?
- How AI is changing the game for the tax and accounting industry
- Case studies: what can ChatGPT actually do?
- Accountability and bias, and the limitations of current generative AI
- What’s next for accountants and tax advisers?
- Our strategy for technological innovation and AI
- The Future is Now: Investment in AI
- Footnotes
- Bibliography
Download the eBook to get access to the full content
Introduction
Rapidly evolving technology is taking the world by storm, and businesses in all industries are looking for use cases to take advantage of the new technologies. Accountants are no different, and the use of AI-driven data analytic tools is common in our industry.
In 2023 Generative AI has become known worldwide, especially with the release of tools on the cloud generally available for free. To begin, this paper looks at generative AI through the lens of accountants and tax practitioners.
Next, the paper looks at the industry code of ethics and legislative codes of professional conduct that govern accountants and tax practitioners.
The legislation and regulatory guidance are reviewed through the lens of generative AI so that you are fully informed of their contents and your obligations. There are a number of factors to take into account when entering into cloud arrangements, including how information is transferred between systems and data integrity is maintained.
The paper critically analyses possible use cases for accountants, and then reviews a series of case studies that are designed to provide you, an accountant in business or in public practice, with an understanding of what a current generative AI tool can do. This may assist in your decision-making about how you may wish to progress in your business with generative AI.
Once we review what generative AI can and cannot do, the paper reverts back to first principles. What do you need to do to future-proof yourself as a trusted advisor to your employer or your clients?
Glossary
Artificial Intelligence (AI) is a field of science that is used to describe systems that mimic natural intelligence displayed by humans, such as recognising voices and understanding language.
AI systems are designed to solve complex problems, and they can be rule-based, logic-based, or machine learning-based.
Machine learning, large language models, predictive analytics and generative AI are all subsets of artificial intelligence.
Machine Learning (ML) | Application of AI that systems use to learn based on the data and training humans give them, both explicitly and implicitly. |
Generative AI | Uses the patterns it has learned to create something entirely new. |
Large Language Models (LLM) | Application of generative AI principles to language, allowing computers to understand and generate human-like text-based inputs and responses. |