Artificial intelligence and robotics
How lawyers can benefit from ChatGPT and other big language models disrupting the legal industry
ChatGPT and other conversational AI built with large language models could fundamentally change how the legal industry operates, giving early adopters a huge advantage.
For example, conversational AI can automate routine tasks such as reviewing documents and analyzing contracts for lawyers, helping them to conduct legal research and writing more efficiently. Law firms can also use it to create blog posts and social media content for marketing purposes.
“If you’re a lawyer and you don’t use this stuff, your opponents do. They’ll do a better job than you,” says Noah Weisberg, entrepreneur, former corporate attorney and co-author of the book. Artificial intelligence for lawyers. “For most people, that would give them an advantage, and I don’t know why you wouldn’t have that advantage.”
OpenAI’s generative primitive transformer models, called GPT, are the most well-known among large language paradigms, which can understand, process, and respond to human language. Google also recently launched a preview beta of its AI Bard conversational tool. In addition, Microsoft’s BingAI search engine has OpenAI’s GPT technology built into it. In February, Meta introduced LLaMA, which stands for Large Language Model Meta AI. It is available to academics, policy makers, and others who apply for a noncommercial license.
More large language models are in the pipeline, says Daniel Martin Katz, a law professor at the Chicago-Kent School of the Illinois Institute of Technology. He says we are in the early stages of this technology. Overall, the market for AI devices and services is expected to grow to $90 billion by 2025, up from $36 billion in 2020. According to UBS.
“There has been a fundamental increase in the capabilities of these tools, these large language paradigms, especially with GPT but in general, and that affects the kind of work that lawyers do,” says Katz. “This is important for lawyers because we have technology that is eventually becoming very good at language, and that was always a challenge.”
a LexisNexis Poll A study from March revealed that 57% of consumers are familiar with generative AI tools like ChatGPT. But awareness was significantly higher among lawyers at 86%, half of whom had already used it in their work or were planning to do so.
According to the survey, “84% believe that generative AI tools will increase the efficiency of lawyers, paralegals, or law clerks.” In addition, “it is not only the practice of law, but also the practice of law.” 61% of lawyers and 44% of law students also believe that generative AI will change law schools and the way law is taught and studied.
Conversational AI is a tool that will create new opportunities and free lawyers from tedious tasks, says Weissberg, who founded Kira Systems, a legal technology startup that uses AI to help lawyers review contracts, in 2011 in Toronto. He sold the company in 2021 to Chicago-based Litera. He later became CEO of Zuza, a contract analysis tool.
“There are some issues that we’ve been working on for a long time that you can see were resolved very quickly,” Weisberg says. “But pulling the data out of the contracts wasn’t right about some things.”
Weisberg says that ChatGPT is good at solving some very difficult problems. But the technology still has trouble extracting the data from the contracts accurately, and it still needs someone to double-check the information.
He says GPT-4 is not ready as a standalone approach to contract analysis if expected accuracy is important. “I have no idea when it’s going to be right or wrong, which is kind of worrying,” Weissberg adds.
not a substitute
With the increasing capabilities of large language models and neural network technology, there is growing concern about the future of lawyers. But AI chatbots like ChatGPT won’t cause many lawyers to be laid off, says Darren Orzechowski, partner at Allen & Overy. He is the co-chair of the technology practice at the company based in Silicon Valley.
“It’s a tool for efficiency, and I don’t think it replaces people,” Orzechowski says.
Orzechowski adds that the technology may open new fields for lawyers and specializations. For example, he points out that technological advances such as desktop computers have changed the way people work but have not significantly reduced the number of lawyers in the industry. However, conversational AI tools will change how lawyers practice law.
in FebruaryAllen & Overy has integrated Harvey, an artificial intelligence platform built on a version of OpenAI’s GPT technology that focuses on legal work, into its global practice. Harvey will enable the firm’s more than 3,500 attorneys across 43 offices operating in multiple languages to access and create “legal content with unparalleled efficiency, quality, and intelligence,” according to a press release. During Harvey’s trial run, the firm’s attorneys raised nearly 40,000 inquiries regarding their clients’ day-to-day work in contract analysis, due diligence, and regulatory compliance.
Harvey.AI is not the only startup using AI and large language models in the legal field. Other companies include Logikcull, LawGeex and DISCO. In early March, Logikcull launched a suite of AI features and ChatGPT integrations that aim to reduce discovery time by up to 90 percent, says Andy Wilson, CEO and founder of the company.
Logikcull, with approximately $40 million in venture capital funding, has built a command-line tool for document search using ChatGPT. Wilson says it allows users to control all kinds of actions on a sheet of paper such as “summarize this document” or “translate into English.”
“It’s just a massive legal productivity improvement,” says Wilson.
Logikcull customers spend more than 65,000 hours per month using its software to search through the data they upload, such as Slack messages, emails, and audio files. Wilson adds that they then read it to find relevance. What conversational AI in these large language models now enables them to do is do this reading to them at scale.
“The biggest problem the legal system will have is that its business model is based on inefficiency and billable hours,” says Wilson. “For people who use fax machines and print their emails, that would be amazing.”
Technology is getting smarter every day. The more people interact with technology, the smarter it gets in a process called machine learning. For example, in March GPT-4 passed the bar exam with a score of 75, which is above the average of 68% and puts it in the 90th percentile, says Katz. In a previous test, the GPT-3 test had a score of 50 and only passed the two multiple-choice parts of the test, putting it in the 10th percentile.
In just four years, Katz adds, GPT has advanced from 0% on the Multistate Bar test for GPT-2 to its score of 75. Katz writes, “Large language paradigms can meet the standards applicable to human attorneys in nearly all jurisdictions in the United States.” By tackling complex tasks that require deep legal knowledge, reading comprehension and writing ability. in his paper Posted on March 15th.
With conversational AI, lawyers must also consider the ethical obligations of attorney-client privilege. Renee McDonald Hutchins, dean of the University of Maryland Francis King Carey School of Law, says ChatGPT keeps all the inquiries it receives. Ultimately, it assumes that law firms will likely build their own internal systems using their own data centers.
Hutchins adds that AI-powered large-language models are just as good as the data they are fed. She says that some of this data can have embedded errors. And if the dataset on which the AI is trained is biased and has privileges built into it, the AI will recreate those results.
“Sometimes we don’t want the law to replicate existing hierarchies,” says Hutchins. “If we rely on it too much, we risk repeating oppression in a way we don’t want.”
For example, suppose a judge uses conversational AI and large language models as tools in the criminal system for sentencing recommendations that are based on arrest records. In that case, they may be repeating racial prejudice, Hutchins says, because police arrest blacks and blacks at a higher rate than whites.
“(They) can be very powerful tools,” says Hutchins. “We have to be very attentive to the restrictions[it imposes].”
Another potential issue is that ChatGPT and other conversational AI and large language models may encounter regulatory barriers if someone other than a lawyer is using ChatGPT to provide legal advice and assistance to laypeople; Hutchins warns that this could be considered an unauthorized legal practice.
“We need to make sure we graduate students who are knowledgeable and skilled in technological advances, but also bear in mind the need for human experts to guide clients through the legal system,” says Hutchins.
Laura Lorick is a publisher Silicon Hills Newsis a regional technology publication headquartered in Austin, Texas. She is also a host Invoice ideas Podcast. Previously, she was a senior writer for San Antonio Express NewsSenior writer on Interactive Week MagazineTechnology writer and columnist for The sentinel sun.