In October 2016, the entire legal world exploded with the news that an algorithm had been created that guessed the decisions of the Strasbourg Court with a probability of up to 79%. The algorithm's creators, scientists from the U.S. and Great Britain, are confident that it will be in demand over time when selecting and sorting cases by importance.
What is it: a scientific experiment or the beginning of the LegalTech-revolution?
These days, information technology is developing exponentially, which applies not only to technology but also to software capabilities. As a result, the legal profession, like others, is on the verge of significant changes. For example, when searching for information, computer services are already taking over the function of reviewing a large number of documents that used to be performed by lawyers. However, these services are also ready to replace humans in other areas of legal activity, from drafting legal documents to predicting the outcome of a lawsuit. In doing so, intelligent machines will become increasingly efficient regarding their functions and costs. And unlike humans, they can work nonstop, without sleep or rest. This continuous technological improvement distinguishes models based on artificial intelligence from previous models of development in the field of jurisprudence. This difference defines the central phenomenon that lawyers will face in the coming decades. These changes are evident in every aspect of the legal field: law enforcement, legal science, and education. New technologies will significantly change the perception of the legal profession; in doing so, many lawyers will go bankrupt, and others will benefit. Two categories of lawyers could particularly benefit from technology. First, it is the most prominent members of the profession who will become even more recognizable and be able to expand their influence. Second are those lawyers who would be able to change the organization of their work and take advantage of the low-cost components provided by machines. Such lawyers would be able to expand the market available to them at the expense of middle-class clients and small businesses. The technology could, at least at first, not very significantly, affect other categories of lawyers, both positively and negatively affecting their functions. First, trial lawyers will retain their highly lucrative niche because computers will not be able to make courtroom speeches for the foreseeable future. However, computers will probably reduce the number of debates by offering options for reconciling the parties. Second, legal advisers persuading clients to take action will not lose their importance since machines cannot create emotional connections with clients. But as for the rank-and-file members of the profession who draw up standard wills and contracts and check documentation - their fate is much sadder since computers can perform most of these routine tasks. Therefore, the development of technology that benefits clients will be ambiguous for lawyers.
Areas of law that will be primarily affected by artificial intelligence
New information technology has already changed some of the professional functions of a lawyer, such as searching for information on the circumstances of a case. Other parts remain almost unaffected, such as preparing case files. There is, however, cause to believe that they will also change because machine intelligence is already demonstrating skills similar to those of lawyers in closely related fields such as journalism. Experts highlight the following five areas that will be dramatically changed by machine intelligence shortly:
- searching for information on the circumstances of a case,
- search for precedents,
- drawing up documents,
- preparation of case files,
- predicting the outcome of a case.
1. Case Information Retrieval
Computer technology is most widely used in the case of information retrieval, perhaps because this type of activity is the application of general machine search techniques to databases of legal documents. In its simplest form, electronic searching, or e-search, is a process in which a computer searches a database for human-specified keywords. However, keyword searching is a relatively simple tool. It can produce both too narrow and too broad a set of results since the specified keywords may be absent in relevant documents and present in unsuitable ones. Indeed, this method is imperfect. However, human errors and inaccuracies are also allowed to occur. The quality of reviewing an array of documents is reduced by fatigue, monotony, and other weaknesses not common to machines. As a result, some courts have endorsed its use as a search tool to help improve decision-making, noting that its cost and performance are at least as reasonable as traditional methods. As with other information technology, computer-assisted search methods will become increasingly accurate, cheaper, and thus more widespread.
2. Searching for precedents
For centuries, lawyers have been searching and selecting precedents to build case strategies. Machine intelligence will not only be able to do this work for lawyers, but it will also be able to do it more efficiently. The first experiments in this direction came in the mid-1960s when the Ohio State Bar Association tried to create an electronic system for finding court decisions. It formed the basis of the Lexis system, which appeared in 1974. Westlaw came soon after, but its functionality was limited because it did not produce the full text of the judgment. Lexis was also flawed because it contained an incomplete database of precedents. These shortcomings were later corrected, and Westlaw and Lexis are now the benchmarks for legal search software. Such developments have already become an essential element of legal practice, replacing less efficient ways of searching through collections of precedents. The example of the Watson robot demonstrates one crucial innovation: the shift from using keywords to semantic search. With the development of semantic search, this limitation will be removed. Semantic search will allow professionals to formulate queries to the machine in natural language, and the computer will also respond in natural language without reducing the relevance of the information. Already, we are seeing legal software more accurately structure the data that makes up the content of a lawsuit, focusing more carefully on the legal aspects. Legal cases are not equal in their precedential relevance: it depends on the weight that precedent acquires over time and on the opinion of the court and the judge who construes precedent as more or less persuasive. Significance also depends on the reasoning in which the jurist uses the precedent and on the court and judge to whom he presents the case. Thus, even if a lawyer has found a precedent utilizing a computer, he relies only on his judgment in using it. But machine intelligence can also make judgments about the significance of precedent. Machine intelligence will not only find a precedent, but it will also prompt a lawyer to use it since most lawyers can neither comprehensively assess its strength nor retain all possible precedents in their memory. The timing of these technologies depends on the general progress of search engines, as the field of application is much broader than the legal services market. The first stage, the refinement of semantic search to assess the strength and applicability of precedent, is believed to come within the next ten or fifteen years.
3. Documents as Forms
Legal document forms are not a new phenomenon. Since the Middle Ages, lawyers have used legal claim templates to help keep costs down. However, machine intelligence is set to revolutionize how legal documents are written. The most obvious is changing the document template to fit the individual situation. However, as technology and software evolve, computer-generated forms will become increasingly common in legal practice. Some new companies are already using machine intelligence to create various documents. In the future, documentation will also improve in the direction of being more connected to the outcomes of lawsuits. With an advanced system of linked data, the machine can correlate the details of contracts with all the court decisions in which they appear, thus providing the basis for continuous improvement of legal forms. Of course, lawyers will first have to scrutinize the documents created by the machine. But even at this stage, the time savings can be very significant. For example, even today, the average time to prepare a package of documents has been reduced from 20-40 hours to just a couple of hours. When designing a large package of documents, the time to compile them is reduced from days and weeks to hours.
4. Documents as case materials
Artificial intelligence won't stop at automating form drafting. Legal forms are easier to automate because they often require formal data entry. At first, such documents will be rough sketches that need additions and fine-tuning. Nevertheless, computer-generated outlines can make a lawyer's work much more manageable, comparable to the result of a paralegal without a law degree. Eventually, such programs will be able to create drafts of documents, as well as connect to legal search programs and extract data from there. As with legal data retrieval, significant progress in developing such programs is expected over the next fifteen years. During that time, programs will probably learn how to create draft documents. At least for restricted applications, it will take another decade or two to produce entirely produced documents.
5. Legal analytics
Predictive analytics is a new field of science that analyzes data to make predictions. Computational power allows information to be collected and structured. Patterns can then be looked for in that data. Machine learning helps analyze patterns and build models. With these models, it is possible to predict the development of situations that have not yet occurred based on the available data. Predictive analytics is now very much in demand in business as well. One of the most important trends of the last decade is the use of "big data" in decision-making. Legal information includes factual patterns, precedents, and the results of court cases. In particular, legal analytics can predict an issue's outcome, enabling a lawyer to assess the work of a lawsuit. To be sure, lawyers implicitly evaluate a lawsuit's prospects when they advise clients to file or drop a case. But their advice is based on intuition and limited to their direct or indirect experience. The advantage of predictive analytics is that it offers a mechanism both for accessing a wealth of information and for systematically processing that information to predict the outcome of a particular case. Legal analytics is a thing of the near future; scientists and companies are already working to implement it. For example, political scientists have created a U.S. Supreme Court decision-making model based on past decisions that are more accurate in predicting the outcomes of lawsuits than that court's team of experts. Perhaps predictive analytics will still be imperfect, capable of making only approximate predictions. However, it will displace some lawyers whose predictions will be even less accurate. Or rather, legal analytics will still leave some role for lawyers. Specialist judgment can lend weight to predictions made by machine intelligence, even if those predictions themselves are more accurate than the specialist's predictions. Over time, however, legal analytics will diminish the value of lawyers' judgments, at least in some areas.
Future trends
Those areas of legal practice closely related to information retrieval, namely, seeking information on the circumstances of a case, looking for precedents, creating documents like forms and case files, and making predictions about case outcomes constitute a significant body of work for many legal professions. As a result, those lawyers who handle the common elements of such services will experience increasing competition from machines. Moreover, as many aspects of the law become commoditized through artificial intelligence, information technology will make many legal services more transparent, which will lead to further market power for lawyers. The most obvious example of this transparency is the ability of consumers to compare the cost of legal services. But there will also be new platforms that allow him to compare the quality of legal work. Already, some companies compare the effectiveness of individual lawyers based on the data available to them. Some of these innovations will be in the traditional activities of a lawyer, such as creating new forms of ordinary transactions or generating unconventional arguments. In addition, law firms can use computers for ancillary work, thereby reducing their personnel costs. Third, for a range of actual transactions and court procedures, even a tiny improvement in the outcome of a case will result in a favorable outcome for clients. If even artificial intelligence makes great strides in service delivery, the involvement of human intelligence can provide the best work. Machine intelligence can also help lawyers increase the number of services they provide in the low-price sector through improved skills or organizational aspects. There is currently a great need for legal services for low- and middle-income people who cannot afford to hire a lawyer at current rates—these range from advising small businesses to drafting prenuptial agreements. Lawyers can draw up the necessary forms using computers, thereby reducing the cost of services and making them more widely available. In other areas of legal work, computers may have less impact. For example, machines will not be able to appear in court and thus will not displace those specializing in lawyering. However, the indirect effect on this area will also affect trial lawyers. As predictions about case outcomes become more accurate, the number of court cases will decrease, as parties are more likely to come to a pre-trial agreement. Legal analysis is not all that lawyers do. They also establish a trusting relationship with their clients, thereby building clients' long-term interest in legal services, even if the client's feelings do not make them aware of that interest. Computers are unlikely to be able to establish such connections and thus will not significantly affect this critical aspect of the lawyer-client relationship. Thus, the overall effect of introducing machine intelligence into the legal profession will be ambiguous. Particularly problematic will be the situation of mid-level and below-average lawyers with no narrow specialization. For consumers, the development of artificial intelligence will be highly beneficial since it will reduce prices and increase the transparency of the legal profession. The middle and lower classes will significantly benefit, as they can receive legal services at affordable prices.
Help in the Work of Lawyers
Members of the legal profession recognize and even actively use computer technology as an aid. To encourage this development, the American Bar Association (ABA) has revised its rules to permit, and in some cases even require, the use of artificial intelligence. In doing so, the ABA has defined the position of artificial intelligence as a tool like a computer or even as an assistant without legal training. Under this understanding, lawyers are mandated to use computer technology if it is necessary to provide skilled services or desirable if it helps improve the quality of services offered. Thus, mastery of technology related to the provision of legal services, including machine technology, becomes a direct responsibility of every qualified lawyer and a significant characteristic of their skill level.
The Increasing Role of Artificial Intelligence in Legal Services
The market for electronic legal services is at a relatively early but crucial stage in its development, which will lead to artificial intelligence playing an increasingly important role in five areas of legal work: case information research, case law research, document drafting, case file preparation, and predictive analytics. Over time, computers will be able to provide these services faster and more efficiently than many lawyers. Thus, ultimately, the development of machine intelligence will break the monopoly of lawyers, benefiting consumers and society as legal services become more transparent and accessible, as will justice in general.