Get Started
Russ Elmer with light grey background

ARTICLE | June 10, 2024 | VOICES

AI will transform legal services delivery 

Artificial intelligence will help attorneys focus on work that matters 
By Russ Elmer, General Counsel, ServiceNow, Workflow contributor

Don’t worry, lawyers. Artificial intelligence (AI) isn’t going to take your jobs—unless you don’t use it.  

Augmented by AI, legal teams can accomplish more. They can avoid getting bogged down in repetitive, time-consuming tasks and focus on work that matters. Not only will attorneys work more efficiently, but they’ll work smarter too, as AI suggests possible answers to common issues.  

Attorneys who make effective use of AI can gain a significant edge. The key to realizing these benefits is to understand how AI can help and how to implement it effectively. Here are some considerations to keep in mind. 

Related

Building an Enterprise AI Governance Plan

Read now

Many in-house legal teams spend a great deal of time fielding routine questions from employees. They want to know how the company’s stock plan works or what the policy governing paid leaves of absence is. They may even need help resetting the password for their online retirement account—something I get asked about regularly, even as ServiceNow’s general counsel. My team prides itself on being a service function—we want to serve our colleagues, even if that means answering questions that are already addressed in our extensive library of knowledge base articles. But fielding these inquiries isn’t the best use of our time. We’re forced to choose between being helpful and staying focused on work that more directly creates value for the company. 

AI can free legal teams from this dilemma. It can learn about the most common questions people are asking the legal team and serve up personalized answers to them. Rather than referring to a static FAQ, people can interact with a chatbot that can anticipate their needs based on role, time of year, and a variety of other factors. Not only does this help people get better answers faster, but it’s also an efficiency booster for the legal team because the questions never come to them in the first place.

Many of the tasks in-house legal teams face require a great deal of care, but the work is mostly standard. Take nondisclosure agreements (NDAs), for example. They can be very long and complicated. AI has the potential to help immensely, allowing a company’s attorneys to examine an NDA provided by a customer or other party, compare it to the company’s standard agreement, identify issues, and suggest revisions. That's not taking away work from an attorney; it's helping an attorney get the work done faster. 

That can lead to two things. First, it speeds up the routine work to clear room in the attorney’s day to tackle the more complex, time-consuming projects. Second, it helps the attorney get their work done faster so they can spend time with friends and family, or focus on a hobby, or simply relax and recharge. Burnout and exhaustion aren’t badges of honor.

For generative AI to be useful to lawyers, whether they are part of an in-house legal team or working at a firm, it must be safe and reliable. There’s no room for hallucinations or other misfires. And there can be no risk of exposing sensitive information.  

That’s why the use cases my team has explored—and those I discuss with ServiceNow customers—won’t leverage publicly available large language models (LLMs) like ChatGPT. Instead, we’ll be using our own LLM, trained on our own data, including information in our knowledge base; real questions we’ve fielded in the past; actual contracts that we have reviewed, amended, and approved; and so on. This will help us prevent common risks regarding privacy and data protection, copyright, and hallucinations. And it will help ensure that the information our LLM produces is tailored to our needs.  

Legal teams that embrace AI’s potential will have a sizable advantage over those who hesitate.

That’s the ultimate vision. In the shorter term, however, we’re experimenting. That means starting smaller, with less important questions and a more public dataset. We’re keeping a human in the loop at all times and not relying solely on the model’s output. Most importantly, we’re getting started—and we’re learning. 

As we put AI to work, we’re guided by three simple principles. 

First, don't wait for it to be perfect—just get started.   

Second, look for quick wins. These could be the rote tasks that cost attorneys a lot of time and pull them away from high-value work. Begin to use AI bit by bit to help them with these jobs, taking stock of what works and what doesn't. As the most promising use cases become clear, ramp up slowly and build on prior success.   

And third, find your champions. People who enjoy experimenting with a new tool or who quickly see improvements in their day-to-day routines are perfectly positioned to become evangelists for AI’s potential.   

AI has the power to take attorneys’ jobs and make them better. Legal teams that embrace its potential will have a sizeable advantage over those who hesitate. 

Trending articles

Related articles

Impact AI: 2024 workforce skills forecast
ARTICLE
Impact AI: 2024 workforce skills forecast

Emerging technologies will disrupt entire industries while creating new opportunities for people with the right skills. New research predicts how labor markets will evolve as AI gets to work. 

The new power couple: AI and people
ARTICLE
The new power couple: AI and people

With the right skills, employees and organizations can thrive in the age of AI

Governing AI’s future
ARTICLE
Governing AI’s future

Executives need to incorporate well-thought-out governance into every AI deployment 

AI in the enterprise
ARTICLE
AI in the enterprise

How artificial intelligence is transforming work 

Author

Author Russ Elmer
Russ Elmer is ServiceNow’s general counsel 
Articles by Russ Elmer