ad
ad
Topview AI logo

Generative AI│What’s the potential for tax function transformation?

Science & Technology


Introduction

The impact of artificial intelligence (AI) on individuals, organizations, and society continues to be a hotly debated topic. According to a recent Global Tech survey, 57% of business leaders believe that AI will help them achieve their business goals within the next three years. However, many organizations still face unanswered questions about how to effectively prioritize their AI transformation efforts.

The Transformative Potential of LLM Technology

Large Language Model (LLM) technology possesses transformative potential for two main reasons:

  1. Enhanced Productivity: LLMs can significantly boost productivity by generating new content and efficiently searching and summarizing existing material, outperforming any previous technology.

  2. User Accessibility: These models allow non-technical users to engage with AI through natural language, making the technology more approachable.

Despite the excitement surrounding generative AI, it is important to note its limitations. It is known to be prone to errors and can produce "hallucinations," where it fabricates answers to fill the gaps in legitimate content. Data privacy is another growing concern; when users share data with tools like ChatGPT, they risk exposing sensitive information to third parties for model training, with the possibility of that information becoming accessible to other users.

Key Applications of Generative AI

The applications of this technology can be categorized into four broad areas:

  1. Content Generation for Tax Teams: Generative AI can assist in creating research memos, summarizing technical positions, and supporting complex decision-making processes that involve extensive document reviews.

  2. Information Review: The ability to analyze vast amounts of text allows generative AI to distill essential information quickly. It can be utilized in fact-checking processes and to identify discrepancies in documentation relevant to tax aspects.

  3. Smart Conversations or Chatbots: Traditional chatbots have often appeared clumsy in interaction. However, LLM-powered virtual assistants can offer a more human-like conversational experience.

  4. Code Generation: Generative AI enables tax teams to interact with databases using natural language queries, eliminating the need for writing code.

Strategic Considerations for Adoption

Organizations should consider how generative AI can enhance the effectiveness of their tax functions. Here are some crucial questions to reflect upon:

  • Does your organization have a defined stance on the utilization of generative AI?
  • What risks and opportunities do you perceive in adopting generative AI, both for the business and for the tax function?
  • In what ways can generative AI enable the tax function to adopt a more strategic role and provide greater value to the business?

KPMG's Tax Technology experts are available to help you navigate these questions and explore the possibilities that generative AI can unlock for your tax function.


Keywords

  • Generative AI
  • Tax function
  • Large Language Models (LLM)
  • Productivity
  • User accessibility
  • Content generation
  • Information review
  • Chatbots
  • Data privacy
  • Code generation

FAQ

What is Generative AI?
Generative AI refers to artificial intelligence systems that can create new content, such as text, images, or code, based on the data they have trained on.

How can Generative AI enhance tax functions?
Generative AI can streamline tasks like content generation, information review, and enhance interaction through more effective chatbots and code generation.

What are the risks associated with using Generative AI?
Risks include the potential for errors or hallucinations in content generated, as well as concerns around data privacy when sharing sensitive information with AI tools.

Who can help businesses implement Generative AI in their tax functions?
KPMG's Tax Technology experts are equipped to assist organizations in understanding and implementing generative AI capabilities tailored to their needs.