
And we already know that with a few clicks, a computer legal research program can quickly turn up virtually all the relevant law and secondary sources needed in a particular situation-provided the researcher frames the search well-working from vast amounts of data. Additionally, machine learning is on its way to dominating the discovery process. We already are seeing AI review contracts and other kinds of transactional “boilerplate” legal documents that vary little from transaction to transaction. The ROSS program, based on Watson technology, answered legal questions. IBM’s Jeopardy-Game-winning Watson program is, perhaps, the most famous example of a QA system. Question answering (QA) systems search large text collections to locate documents, short phrases, or sentences that directly answer a user’s question. In a citation network, the connected objects may be legal cases or statutory provisions. Network diagrams graph the relations between objects and can assist in making legal information retrieval smarter. Machine learning (ML) refers to computer programs that use statistical means to induce or “learn” models from data with which they can classify a document as relevant to a claim in e-discovery or even predict an outcome for a new case. Legal text analytic tools employ machine learning, network diagrams of citations, and question answering. Legal text analytics refers to techniques for automatically extracting information from archives of legal documents including case decisions and statutes. Recent developments in AI and in legal text analytics have led to some new tools for legal research, many of which are now familiar to most lawyers and legal educators. AI and Law, in particular, focuses on modeling legal reasoning behaviors. AI is a subarea of computer science in which researchers develop computational models of behaviors that normally require human intelligence. Our goal in this Article is to answer these questions: How far can and do AI capabilities go when it comes to legal writing? Can AI do the kind of sorting, analysis, and text generation required for solving legal problems and writing analytical legal documents? What AI skills will new lawyers need in the not-so-distant future in order to be prepared to enter legal practice?įirst, however, here is a brief introduction to the field of AI and Law.
#Bear writer and alexa full#
Even if AI’s ability to write out a full memo is still in the future, at present AI can contribute fruitfully to the memo writing process. The newest AI can draft a paragraph or two of apparent legal analysis in a sophisticated kind of cutting-and-pasting based on texts in its enormous collection but with no guarantee of legal accuracy.Īs artificial intelligence (AI) capabilities become more and more sophisticated, legal educators and practitioners may have to confront the possibility that in the foreseeable future, AI will be able to write the basic and foundational legal document, the one with which we train our students to first write out legal reasoning, the office memorandum.

For certain types of cases, AI can predict an outcome given a textual description of a case’s facts, but it cannot yet provide an explanation or justification. For a limited variety of legal memoranda, AI can draft a memo automatically.

Having done so, however, they can submit the draft memo to an AI program for suggestions of additional issues or cases to cite. Students still need to read the cases to select and frame legal questions, identify questionable elements, search for cases concerning those elements and analogize them to or distinguish them from the problem scenario, and ultimately, draft the memo. As we will see, AI currently can process a natural language description of a legal question or short scenario and return answers and relevant cases. This Article offers an assessment of the present status of AI and legal writing capabilities and also provides a glimpse into the future: what AI is poised to do in the near future and which attorney writing activities seem beyond the reach of AI.
