Practical Takeaways on Generative AI from Everlaw Summit
If you’ve been paying attention to the legal space in the past two years, it’s likely no surprise that the technology on everyone’s minds is generative AI. But now that it’s firmly integrated into the lexicon of legal professionals, attention has gone past the nuts and bolts of how it works, and turned toward responsibly implementing it in an actionable way.
At Everlaw Summit ‘24, the panels and conversations around this technology focused on the practical, secure ways it can help legal teams transform their work. From erasing inefficiencies to taking on higher case volumes to training young attorneys, there was no shortage of possibilities being discussed. For many organizations, it was clear that the time to act on innovation is now.
This is an exciting time for the legal profession, and Everlaw Summit proved the perfect forum for these discussions around the future of the industry to take place. Attendees had an opportunity to hear from thought leaders and share how they’re thinking of using or already incorporating generative AI into their practice. And while change may not be immediate, thinking about the practical uses and implications of this technology is an important first step.
Takeaways on Generative AI from Everlaw Summit
Here are just a few of the main takeaways surrounding generative AI from Everlaw Summit ‘24.
Getting Started Is Often the Hardest Step
Many of the speakers at Everlaw Summit discussed the fact that getting started is often the most difficult step when it comes to using new technology. Changing familiar habits and processes can be difficult, but even just getting set up with the technology can go far in driving adoption.
“It's not about winning or losing when it comes to generative AI per se,” Mark Abramowitz, a partner at DiCello Levitt, said during a panel on generative AI for plaintiffs firms. “It's about, are you taking that first step right now? And how committed are you going to be to that first step? The fact is you're looking at this technology, and you're using it to help your side. It’s about that adage we always hear of doing more with less.”
This was echoed later in the day during a panel with public sector speakers on the topic of complex litigation. “Getting the public sector up to speed with generative AI is going to be its own challenge,” Missouri Deputy Attorney General Jeremiah Morgan said. “I've heard from a few clients that AI will help attorneys general level up, if you will, when going toe to toe with private law firms. Getting us on board is no small thing, but it has the potential to really make a difference.”
Although introducing generative AI tools in the public sector can present its own challenges, many organizations can get started by doing little things, like creating a secure sandbox for people to experiment in, holding interactive training sessions, and letting people know they shouldn’t be afraid to fail.
Patience Is a Priority
As with any long-term change, patience is required. You have to be patient with your organization as they onboard a new system, patient with yourself as you learn how to use it, and ultimately patient with the technology as it gets up to speed.
“The lawyers are going to be worried about generative AI, and are going to have that fear that the technology is missing something,” Mark Agombar, Director at XBundle, said in a panel on uncovering key evidence in high-profile litigation. “But we have to be responsible in educating them on how this is going to improve accuracy, and not try and blind them with science or anything else like that. So many of the lawyers who are concerned about the imperfections of the technology never pay attention to the imperfections of human review, and where that falls short.”
Mistakes will be made, both on the part of the legal professional and the technology, but if you’re patient and serious about learning how to use it effectively, it will pay dividends later on. Those growing pains are all part of educating yourself in a responsible way. The hard truth about generative AI is that mistakes will happen, but with proper training and guardrails in place, those opportunities for mistakes will be minimized.
Privacy and Confidentiality Are Table Stakes
As AJ mentioned in his keynote on the second day of Summit, privacy and confidentiality should be top of mind for everyone who is considering using generative AI. It’s a core part of Everlaw’s AI principles and values, and pervaded all three days of the conference.
In a profession where sensitive and confidential information abounds, legal organizations should be aware of how their data is being used, and whether it’s retained by generative AI providers. There should be an agreement in place that ensures no data is used to train the generative AI models or retained for any purposes outside of responding to the initial query.
Prioritizing privacy and confidentiality will not only keep your organization’s and your client’s data secure, but it will also prove that you’re leveraging this technology in a responsible way. You should never compromise on the privacy of your data, or work with a provider that can’t guarantee its security.
Making Technology Work for You
With legal technology, and generative AI specifically, it’s all about utilizing it in the right way. If you know how to implement the technology properly, it can be an effective way to become more efficient and work smarter.
“What if, with generative AI, certain parts of your case can be done instantaneously before you're even thinking about it?” Abramowitz said during his panel. “It’s that idea of 80 over 20. It'll get 80% of the work done, so you can really focus on that 20%. Because the fact is these systems aren’t at 100% yet. But if it can do 80% of the work, if it can take my documents and summarize them in a way that an attorney can then take and crunch it down afterwards, that's a big win to me. Even if it only gets me 60% of the way there, that to me is still a win.”
Generative AI allows organizations to increase efficiency and streamline workflows like never before, so you can dedicate attention to the legal work that will help you win.
Others might be thinking of incorporating generative AI even outside of ediscovery or legal work all together, using it for things like Q&A chatbots that can streamline HR processes, onboarding new employees, and more. There are practically unlimited use cases, which is just part of what makes generative AI so exciting.
It’s About Changing, Not Displacing
Generative AI isn’t going to displace attorneys and legal professionals, but it is going to change how they work.
Document review times are going to be greatly reduced, with generative AI being able to comb through document sets and pull out critical information in seconds. Deposition preparation will become much easier, with generative AI generating witness, exhibit, and topic summaries, and identifying any inconsistencies or discrepancies. Developing arguments will benefit from generative AI’s ability to craft arguments, argue counterpoints, or analyze gaps in potential evidence.
As Greg McCullough from Fire Litigation Consulting said, “I've always thought that when I'm using the AI tools, it makes me feel like when Iron Man gets his suit, and now he has these superhuman powers. I'm still operating it, and I'm still controlling it, but I have powers beyond what I'm normally able to do.”
Although much of the conversation around generative AI often falls back into the trope that it will replace humans, the reality is that it will do just as described above: increase efficiency and enhance their work. There is a lot the legal profession stands to gain with this technology, and those who implement it responsibly will have the competitive advantage.
Actionable Ways Generative AI Is Already Being Used
On the Everlaw Summit mainstage, we heard from two people who are actively leveraging generative AI in their firms. Cal Yeaman, a project attorney at Orrick, kicked things off by detailing an interesting use of generative AI in the document review process, with the Coding Suggestions feature of EverlawAI Assistant.
Yeaman used a combination of predictive coding and Everlaw’s generative AI tool, EverlawAI Assistant, during the review of more than 10,000 documents in an attempt to unlock insights more efficiently. It yielded usable results in a matter of minutes, which is incredible considering the fact it would’ve taken an attorney several days. What’s more, EverlawAI Assistant was consistent across the board, and only missed one document in its search that an attorney later deemed relevant.
“Even if you’re being inefficient with generative AI, and have to run a search four or five times before you know what you’re looking for, that still costs less than half of what it would in billable hours for an attorney to go through the documents themselves, which is a massive savings to the client,” Yeaman said.
He recommends experimenting with generative AI on closed, complex matters to get comfortable before moving on to live matters. Get to know your document populations first so you know what to look for. And, ultimately, try to break it.
“Give the generative AI model different things to analyze in different ways and try to break it. See how it works, how it doesn’t work, where it stands up and where it falls down. Then try to figure out why,” he said.
After Yeaman spoke, Jen Jackson, who is a Senior Analyst at Baker Dolinko & Schwartz, took the stage to detail her use cases, leveraging Storybuilder and the Writing Assistant feature of EverlawAI Assistant to assist attorneys on a couple of different matters.
The first involved a case that had been on pause for over a year and was now active once more. The case law was more than 150 cases, and the attorneys wanted a sort of study guide to summarize all of that information. While Jackson first thought about uploading everything into ChatGPT, her concerns over hallucinations caused her to think EverlawAI Assistant. She was able to quickly get the three most relevant cases and pass them over to the attorneys handling the case.
For another matter, the firm had been served with 10 RFPs, which included over 120 pages of requests and 500 items. Instead of handing the task off to a junior attorney and wasting both time and billable hours, Jackson instead uploaded the documents to EverlawAI Assistant, and got a condensed two page summary of the tasks in a matter of minutes.
To those just getting started with using generative AI, Jackson said, “There seems to be a lot of tedium involved in the legal industry. And while that can often be necessary, it’s not typically impactful. My advice is that when those tedious tasks come up, generative AI should come to mind. You can ask the Writing Assistant anything you want, and see what it comes back with. We all know that a blank screen is a much more difficult starting place than a strong first draft.”
The Next Iteration of Legal Technology
Generative AI is ultimately the next iteration of legal technology you’ve already been using. From case management systems, to TAR, to machine learning, and now generative AI, this technology can help you do more with less, but isn’t meant to replace your role or do your job for you.
This next generation of legal work will be about leveraging this technology to increase efficiency, lower costs, and continue providing top-shelf legal services to clients. While that might not sound groundbreaking, generative AI will help legal professionals accomplish these goals in ways they haven’t been able to before.
Now that Everlaw Summit has passed, it’s time to use the knowledge and skills gained over your three days to take action and begin implementing generative AI into your workflows. Talk with your team or generative AI provider to see how you can start the conversation around taking the next step.
