When AI innovator OpenAI released ChatGPT in November of 2022, it sent shockwaves through the world, from tech leaders to academia. Quickly, stories surfaced of people using ChatGPT to “cheat,” from students turning in papers written by the app to lawyers filing briefs that cited cases that didn’t exist.
Beyond the initial splash and the mania both for and against this new tool, the legal profession is finding both practical uses and reasons for concern when it comes to generative AI. The only thing that’s certain is that its impact will be profound.
What is generative AI?
Generative AI uses two processes to create original content based on conversational prompts by a user. Natural Language Processing (NLP) enables computers to comprehend, and then generate, natural human language; and Large Language Models (LLMs) allow an AI tool to search massive quantities of data and surface relevant information.
It can be a powerful time-saving tool, as evidenced by a recent MIT-Sloan study that found generative AI could boost worker productivity by 40%.
What are the drawbacks?
The primary challenge with generative AI is that it’s only as good as the data it has available and the precision with which the LLM being employed has been trained. In legal work, attorneys are required to provide correct information in framing their cases, which means they need to carefully check the results they get from a generative AI tool. When they don’t, they can easily provide incorrect citations or draw fanciful conclusions, impacting their clients’ cases.
Some attorneys might also use generative AI to save themselves time but not pass those savings on to their clients, which would be a form of theft. In most cases, the use of AI as a time-saving technology should be disclosed.
What are the advantages?
In general, AI holds tremendous potential to make skilled attorneys more effective by assisting them in research, drafting, and other time-consuming tasks. Lexis+, for example, is a widely used generative AI tool that helps attorneys research, draft, summarize, and analyze legal texts.
Generative AI also holds the potential to help those who lack access to legal services, which is a persistent problem. The American Bar Association estimates that Americans don’t receive adequate help for 92% of their legal needs. And in rural areas, those in need of legal services often struggle to meet with lawyers and access the court system.
As generative AI applications trained on high quality legal data become generally available, those applications may enable more Americans to effectively represent themselves, expanding access to the justice system.
The upshot
Generative AI can be a valuable addition to any attorney’s toolkit, but lawyers need to use it ethically and responsibly, and with careful regard to client confidentiality. For the general public, AI tools aren’t yet able to provide the level of quality required by the courts. So, while generative AI applications can be used for personal research, briefs, filings, and other legal work should still be completed by a competent, licensed attorney.
