RESEARCH QUESTIONS:
The researcher has answered three questions in the course of this project:
- What is Artificial Intelligence?
- What is the relationship between Artificial Intelligence and Law?
- Can Software replace Lawyers?
SOURCES OF DATA:
Secondary sources of data such as articles, books, reports and cases have been used.
MODE OF CITATION:
A uniform mode of citation has been followed throughout the project.
What is Artificial Intelligence?
In spite of being widely applied to various other professions, computer technology has had little relevance in the legal profession, primarily due to misconceptions that exist between legal professionals and computer specialists of each others profession. However, development of Artificial Intelligence, [hereinafter “AI”] a technological breakthrough, in the past decade or so, may very well change the face of the legal profession. AI is a branch of computer science and programming and is defined as:
“The theory and development of computer systems able to perform tasks normally requiring human intelligence, such as decision making and speech recognition”
The vision of the AI researchers is that no longer will a computer be delegated back-office tasks such as drafting or billing etc.; instead it will now perform the tasks of a lawyer such as legal reasoning, providing legal advice and solving cases. After successful application of AI technology to various fields, (e.g. geology, scientific research, medicine, programming, game playing) computer experts turned to the legal profession to test their new discovery.
AI and Law
The initial twenty five years were spent in developing Legal Knowledge Retrieval Systems [hereinafter “LKRS”], programs that stored vast amounts of legal data, (case reports, case laws, articles, statutes etc.) and could draw it up when fed with the right input of ‘key words’. An improvement on LKRS was suggested by those who believed that by being used only as a storehouse for information; computers were being minimally and impractically utilized. Thus, a system that holds heuristic knowledge, such as that which is present in practitioner’s texts and handbooks, might just be more useful and of far better practical value. Such views paved the way for AI to enter the field of law.
The branch of AI from which Law will benefit is an applied branch and is referred to as Intelligent Knowledge Based Systems and Expert Systems [hereinafter “IKBS” and “ES”]. They are systems that contain representations of knowledge and can be used to in solving problem. ES are a more technologically advanced form of IKBS such that they are capable of performing at a level that an expert in a highly specialized field would be able to.
Given the passage of time, and huge advances in technology, a day might come where a computer with artificial legal intelligence solves all problems. Having laid down the basics of IKBS and ES, the researcher shall now proceed to examine this very possibility of a machine replacing a human brain, replacing a lawyer.
Can Software replace Lawyers?
Ideally, a lawyer is expected to have the following qualities; (1) a thorough knowledge and understanding of law, (2) ability to peruse vast quantities of data, (3) objectivity and clarity of reasoning as well as an unbiased approach (4) speed and efficiency. If these qualities can be achieved on a computer using AI, “‘people seeking routine legal advice might not even think of contacting their lawyer for help. Instead, a computer will use advanced artificial intelligence (AI) technology to answer questions, present options, prepare and file pleadings, and send them on their way,’” as is prophesized by Dr. L. Karl Branting, a lawyer and computer scientist.
The pending backlog of cases in court, the vast corpus of law and the resulting intensity of specialization has made it difficult for judges as well as legal practitioners to keep pace with the constant developments in law. In such a scenario an ES would help a lawyer/judge to tap into the heuristic knowledge of a specialist when he is faced with a problem beyond his knowledge or experience; and also relieve him of the task of researching and referencing a large number of cases and statutes when coming to a decision.
Another function expected of an ES when considering a case is that it should show the client the lines of reasoning used, authorities for all assertions made and conclusions drawn. This transparency of functioning will not only benefit the client, it will also allow the lawyer valuable insight into his reasoning skills. This would enable him to spot flaws or consider other methods of reasoning. An expert system, being essentially inanimate, will be free of the biases inherent to a human being.
It is evident then, that AI technology could vastly improve access to and efficiency of the justice system along with reducing costs of both, courts and clients. People who previously did not have access to the law will be able to avail of it. Also there is the emerging generation of techno – savvy lawyers who actively use AI technology in their practice. A conscious shift by both lawyers and the public in incorporating AI technology in the legal profession based on the factors stated above will be the reason it replaces lawyers.
Law, however, is an open ended, dynamic field. The scope of its content cannot be taken as black or white but consists instead of a number of shades of grey. It is these shades of grey, in the form of the richness and depth of and multiple meanings attributed to the language of law that will pose a challenge to an expert system. While mechanical tasks are easy to perform, it is the ‘common sense’ tasks like natural language comprehension and intuitive reasoning that are the hardest for it to perform.
Also, there are many occasions wherein the Judiciary has stepped beyond the ordinary bounds of law to give a decision. The best example is that of Keshvanada Bharti v. The Union of India. This case laid down the Basic Structure Doctrine [hereinafter “BSD”] when no such concept had existed earlier. It was a landmark decision that was independent of any law in existence.
The very fact that the corpus of law is vast and dynamic and that it also consists of the heuristic knowledge of experts, means, that to successfully codify all of it is a mountainous, if not impossible task. This is because the complex nature of computer programming which also leads to a disparity between what the law actually states and what is actually represented in the software.
As a logical extension of the above argument, it is evident that a large amount of human intervention is involved in developing such a program, such as what precedent to follow, what role does equity and morality play? Hence, when the AI program is dependant on human knowledge and subjectivity from the very start, it is illogical to argue that they would replace what created them.
The interdisciplinary nature of law attributes to it a subjective nature, whose interpretation differs from lawyer to lawyer, thus preventing it from being a set of “cut and dried, ready to apply” rules. Law develops in sync with politics, society and economics to name a few and has a strong grounding in morality and equity.
AI is heavily criticized by legal scholars as it doesn’t give jurisprudence – which deals with the development and applicability of legal theory and legal reasoning – due credit. Without understanding the nature of and context in which a law has been passed the assessment made by an ES would be superficial. There are also practical difficulties, such as money and time that shall be spent on research and development of AI technology.
Therefore the reasoning that software, will one day completely replace lawyers is inherently flawed. It can be unequivocally stated that, under no circumstances can a machine take the place of a human mind. Lawyers, then, have absolutely nothing to fear as they are not in danger of becoming extinct. Their inherent ability to look at a case taking externalities into account, using experience and instinct when faced by heretofore unaddressed legal question, and also the basic fact that instead of being cold machines they can look at a case with a conscience makes them irreplaceable.
However, the researcher would like to counter question both the above schools of thought with a different perspective on this debate. Is it really relevant, or of any consequence to determine whether or not lawyers can be replaced by software? The researcher believes that this question is redundant. Technology has already pervaded the previously ‘orthodox’ legal profession and, in the researcher’s opinion, is here to stay. To raise the question as to the existence of only computer lawyers or human lawyers would be a false dichotomy. Neither extreme can exist. Moreover neither can be compared with the other as their functions and qualities are inherently different.
While the strength of AI lies in its speed and efficiency in handling a large volume of work and being meticulous. precise and objective, it loses out when it is confronted with issues that involve factors inherent to human nature such as the basic intuition required in tackling cases, understanding that law is not independent of equity and morality, including the relevance of jurisprudence, being able to correctly interpret different fact situations and distinguishing or allowing as the case may be.
Law holds as many opportunities for AI, as does AI for law. It involves reasoning using several methods – case – based, analogical etc. Thus, an AI programme will not only need to know about them all, but also how to apply them simultaneously. The law is well documented, thus making it more accessible while developing an AI program. That is not to say that every nuance or nitty gritty of legal knowledge is codified, but the substantial amount that it there is sufficient to give AI a leg up. No legal answer is a clear cut ‘yes’ or ‘no’, but is dependant rather on the goal the lawyers wishes to achieve. “‘This is a feature and not a bug,’ to use a computer scientist’s phrase.” The nature of legal answers only adds a variety and challenge to computer experts. Thus, law provides AI with new horizons to explore.
Due to the intrinsic disparities that exist between law and technology, an ES will never be able to capture the various facets of law. There will always be some complex and subtle question of law that is too elusive and subjective for a computer program to interpret and that is where a lawyer’s practical knowledge in the field will be applicable. The researcher believes though, that there is a possibility of developing an AI program that can competently handle those cases that do not involve complications; where the law can be applied directly or is very cogent.
Thus, it is evident that the fields of law and software technology are essentially separate and cannot be considered as substitutes. However, with the current trends, it would be foolish to assume that law can always remain free of technology and vice versa. Everything undergoes evolution, most fields have incorporated technology without being dominated by it, and so will law. It’s just a matter of research and time.
Conclusion
Through the course of this paper, the researcher has come across two schools of thought. One, which believes that software, will replace lawyers and that this is the need of the hour. And the other which believes that law will never be conquered by a machine. However, the researcher has also realized that though there is no chance of complete replacement, there is a high chance of improvement in the legal field. And inclusion of Software is the essential aid to this improvement.
It has to be understood that both fields have great interdisciplinary potential and rather than indulge in skepticism and misrepresentation of each others intensions, it would prove more efficient and beneficial to cultivate a symbiotic relationship. The question then stops being one of dominance, but of acceptance. In the end it’s not about lawyers being superior to computers or vice versa. It is about working together for a mutual benefit that will ultimately benefit society as a whole.
Bibliography
ARTICLES
-
Anonymous, “AI am the Law”, The Economist Science and Technology Quarterly, March 10, 2005, at http://www.economist.com/PrinterFriendly.cfm?story//id=3714082 (August 20, 2007).
-
Edwina L. Rissland, “Artificial Intelligence and Law: Stepping Stones to a Model of Legal Reasoning”, 99(8) The Yale Law Journal 1957 (1990).
-
American Bar Association, Final Report: Alternate Scenario, 2016: Diary of the Last Lawyer, at http://www.abanet.org/lawfutures/report2002/2016diaryoflastlawyer.html.
-
John Sirman, “Artificial Intelligence and the Law”, 66 Texas Bar Journal 17 (2003).
-
Bruce G. Buchanan and Thomas E. Headrick, “Some Speculation About Artificial Intelligence and Legal Reasoning”, 23(1) Stanford Law Review 40 (1970).
-
Richard E. Susskind, “Expert Systems in Law: A Jurisprudential Approach to Artificial Intelligence and Legal Reasoning”, 49(2) Modern Law Review 168 (1986).
BOOKS
-
The Concise Oxford Dictionary (Della Thompson ed., 9th edn., Delhi: Oxford University Press 1999).
INTERNET
- -
National Law School of India University, Nagarbhavi, Bangalore
Bruce G. Buchanan and Thomas E. Headrick, “Some Speculation About Artificial Intelligence and Legal Reasoning”, 23(1) Stanford Law Review 40, 40 (1970).
The Concise Oxford Dictionary, 70 (Della Thompson ed., 9th edn., Delhi: Oxford University Press 1999).
Richard E. Susskind, “Expert Systems in Law: A Jurisprudential Approach to Artificial Intelligence and Legal Reasoning”, 49(2) Modern Law Review 168, 174 (1986).
DENDRAL, a project developed by Stanford University, is used to determine molecular structure, PROSPECTOR is used for geological surveys and MYCIN, CASNET and CADUCEUS are various AI Projects successfully operating in the field of medicine.
Despite the fact that ES and IKBS are used interchangeably, ES are a type of IKBS that have been constructed with assistance of human experts and are capable of performing at standards higher than that of an expert in the field. The latter might be capable of speech recognition or say image perception that is dependant on knowledge, however does not really require human expertise. The former (ES) on the other hand are almost human in their performance.
John Sirman, “Artificial Intelligence and the Law”, 66 Texas Bar Journal 17 (2003).
Anonymous, “AI am the Law”, The Economist Science and Technology Quarterly, March 10, 2005 at http://www.economist.com/PrinterFriendly.cfm?story//id=3714082 (August 20, 2007).
American Bar Association, Final Report: Alternate Scenario, 2016: Diary of the Last Lawyer, at http://www.abanet.org/lawfutures/report2002/2016diaryoflastlawyer.html (August 12, 2007).
Supra note 1, at 44 – 45.
Edwina L. Rissland, “Artificial Intelligence and Law: Stepping Stones to a Model of Legal Reasoning”, 99(8) The Yale Law Journal 1957, 1959 (1990).
The BSD categorizes certain articles (Secularism, Right to Life, and Right to Equality) of The Constitution of India as essential to its scope and meaning; and thus prevents these articles from being amended.
Supra note 15, at 1961 – 1962.