Does financial advice by AI-based robo-advisors will replace humans? Evidence from ChatGPT and Google Bard

Abstract

The advent of artificial intelligence technologies has transformed various sectors, including taxation. The present study investigates the accuracy and knowledge of taxation advice provided by two leading AI tools, ChatGPT and Google Bard, for Indian taxpayers. Utilizing nine evaluative criteria - Clarity, Accuracy, Precision, Relevance, Depth, Breadth, Logic, Significance, and Fairness, we examine the capabilities of ChatGPT and Google Bard in delivering reliable taxation advice. We used 12 different prompts for each taxation advice from ChatGPT and Google Bard separately. Our respondents are limited to Chartered Accountants having exposure to robo-advisory services in varied capacities. The findings indicate that Google Bard exhibits superior accuracy, while ChatGPT excels in other evaluative categories. Both AI tools, however, demonstrated comparable overall performance. Notably, Chartered Accountants slightly favoured the responses generated by Google Bard concerning the correctness of Indian taxation advice. These findings are significant for three major stakeholders of the Indian Taxation system; the Government, Chartered Accountants, and Assessees, and indicate the need for policy changes in regulating the role of Artificial Intelligence in the Indian financial advice domain. One major limitation of this research is the number of respondents due to the complex and time-intensive nature of the questionnaire. This is the first study in our knowledge that evaluates the authenticity and reliability of financial advice provided by robo-advisors; with a particular focus on tax-saving approaches tailored to the Indian context.

Ähnliche Arbeiten