Modern AI language models like ChatGPT use deep learning techniques to comprehend and produce language that is similar to human speech.[1] It is an excellent tool for assisting lawyers with legal research and analysis since it can comprehend natural language questions and deliver relevant responses. Although there is rising interest in exploring its potential, the usage of AI-powered chatbots in India’s legal sector is still in its infancy. ChatGPT can be especially helpful in legal research, where lawyers frequently have to sort through mountains of data to discover relevant cases and precedents. Lawyers can save a significant amount of time and effort by using ChatGPT to swiftly scan through thousands of legal documents, highlight important details, and provide summaries of relevant cases.


ChatGPT originated through the launch of GPT (Generative Pre-Training Transformer), an AI language model by researchers at OpenAI in the year 2018. GPT was more effective than other language models[2] at that time for tasks like common sense reasoning and reading comprehension.

All other programs process text word by word, unlike ChatGPT which uses “tokens” to process the text. The basic unit used by ChatGPT to calculate text length is the token[3]. They are collections of characters that sometimes, but not always, correspond to words. It is specifically based on character count and may contain punctuation or emojis. This explains why the token count and word count are frequently different.[4]

This means that it uses mathematical operations to predict what token should come next. It doesn’t understand the whole meaning of the text. That is why when it is asked a simple question with a little twist that can not confuse a person with average intellect, it will answer wrong. For example, in its first release, when it was asked this question-

Ramesh’s father has three sons. The name of the first one is Suresh and the name of the second one is Sooraj. What is the name of the third one?

The answer is obviously Ramesh. But ChatGPT was not able to understand this question. However, in the new upgrade, this question is being answered by it correctly as the newer version has become better at processing tokens. The newer version still is unable to fully understand the text let alone understand the law.


Question– ‘A’ from Bihar talked over the phone with ‘B’ from Rajasthan. A agreed to the offer given by B and the terms were to supply furniture to B at 50% advance and the other 50% on delivery. Some dispute arose and now B is refusing to pay the other 50%. what is the place of acceptance of the contract according to the Indian legal context?

ChatGPT– without further specific details about the phone conversation and the intention of the parties, it is difficult to determine the exact place of acceptance in this case. It is advisable to consult with a legal professional or lawyer who can provide guidance based on the specific details and circumstances of the situation.

Human Lawyer– Place of acceptance plays an important role in deciding the jurisdiction of the court to fight the case in the proper court. As a general rule in Law, as per the Indian Contract Act of 1872[5], communication of acceptance is complete in the following way- for the proposer, he becomes bound when the acceptor has posted and the letter is out of his control and for the acceptor, he becomes bound when the proposer receives his letter.[6] Therefore the place of acceptance is from where the letter of acceptance is posted. To understand the concept, following this principle, the answer to the above question is Bihar because ‘A’ is the acceptor here. However, the supreme court ruled[7] that in instantaneous communication like a telephone, the place of acceptance is where the offeror has clearly heard the acceptance of the acceptor.

To sum it up, the answer is Rajasthan as ‘B’ is the offeror here and also because this is a case of instantaneous communication.

One thing can be said for certain. As long as Human Judges are deciding the cases, there is no way an AI can present the whole case in front of that judge. A Human Lawyer is still needed to plead the case in court.


A company based in San Franciso has developed a Robot Lawyer[8] that will tell the defendant what to say in court. Another development worth discussing is Harvey AI[9] which also works on the language model as ChatGPT but with the advancement in which its creators claim that it can understand the law.

Besides these developments, simple technologies have been used for quite a while now; for example, websites like SCC Online[10], Manupatra,[11] etc. have eradicated the need of keeping physical heavy books in the lawyers’ chambers for research purposes. Note-taking apps like Evernote and Notion have helped to make particular notes on cases.

When all these programs are used in combination, it has reduced the workload of lawyers and hence increased the productivity multifold.


One thing for certain is that ChatGPT or similar models will not be able to completely replace humans in any field whatsoever. But they will greatly improve the workflow and efficiency of not only lawyers but also other professionals working in any field. They can, however, replace petty jobs, or will affect the industry in a way where one man can do a job at par with five men therefore that four men could be out of a job.

The change should be embraced rather than feared. To prepare one for this shift, he should at least have a basic knowledge of how these programs work, as stated before, as long as the judges are human, there is a need for human lawyers. Keep yourself updated as a lawyer and there should be no big problems. The following three-step process can help you to prepare for the shift-

  1. Understanding the interpretation- the interpretation of the statutes can change how the law is applied. If a lawyer understands the interpretation, he cannot only improve his pleading in court but also get assured that a machine cannot replace him anytime soon.
  2. Working on communication skills- humans have legal problems, they don’t have an understanding of the law, that is why they approach lawyers for that. Even if a person knows how to use ChatGPT, they can’t fully understand the legal consequences. If a lawyer is good at communication, then it is certain that he will get clients sooner or later.
  3. Working with technology- one should always work on improving efficiency. This is what ChatGPT and other programs are doing. Learn to use them to improve your efficiency in whatever you do.


In conclusion, ChatGPT has the potential to transform the legal sector by offering quick and effective legal research and analysis, but it cannot completely take the place of human lawyers. Machines cannot replace the critical thinking, empathy, and in-depth legal knowledge required by the legal profession. However, the acceptance of AI-powered chatbots like ChatGPT might considerably improve the productivity and efficiency of legal practitioners, freeing them up to concentrate on more difficult legal work by minimizing the time and effort needed for basic tasks. It is crucial to think about the legal and ethical implications and make sure that AI-powered chatbots are used responsibly as the usage of AI in the legal sector continues to develop.

Author(s) Name: Krishna Raj Sharma (Dr. B.R. Ambedkar Law University, Jaipur)


[1] ‘Introducing ChatGPT’ (OpenAI) <> accessed 15 May 2023

[2] Aadya Sharma, ‘The Escalation of ChatGPT: How ChatGPT will exert Influence on the Legal Profession?’ (2023) 3(3) Jus Corpus Law Journal 106

[3] Stanislas Marion, ‘How to use OpenAI GPT tokens?’ (GPT for Work, 24 March 2023) <> accessed 15 May 2023

[4] Ibid

[5] Indian Contract Act, s. 4

[6] Ibid

[7] Bhagwandas Kedia v Girdharilal Parshottamdas (1966) AIR 543

[8] ‘World’s first AI-enabled robot lawyer will tell defendant what to say in upcoming court case’ (The Indian Express, 11 January 2023) <>  accessed 12 May 2023

[9] Kate Rattray, ‘Harvey AI: What We Know So Far’ (Clio Blog) <> accessed 13 May 2023

[10] ‘Home’ (SCC Online) <> accessed 15 May 2023

[11] ‘Home’ (Manupatra) <> accessed 15 May 2023

error: Content is protected !!