Tools like generative AI already have real-world implications in the legal profession. In our previous post, we covered discussions at the Corporate Legal Operations Consortium (CLOC) Global Institute around the disruptive potential of generative AI, and how it will affect the way legal teams scale their services.
Here, we review sessions at the conference around active use cases and immediate steps professionals can take to learn more about the technology.
“Generative AI is truly transformative,” said Eric Ortman, senior director of legal operations at BeiGene, during an open session. “We’re grappling with all that it means.”
Ortman spoke on a panel focusing on the value of generative AI for three practical use cases:
Summarizing key documents, such as when reviewing outside counsel invoices
Using AI as a co-pilot to run key playbooks, for example in contract reviews and performing extraction from historic agreements
Auto-generating legal services requests from plain-text communications, such as email
Building Transparency into Legal Spend
A sea change is coming in how outside counsel legal spend is evaluated and negotiated. With its ability to summarize various tasks and the hours spent on each, generative AI can help in-house teams review invoices for compliance and value – to better understand what they paid for.
“This will change the way we work with outside counsel,” Ortman said. “My expectation is that outside counsel will learn how to embrace the use of generative AI to work more efficiently.”
Aaron Van Nice, ADM Company’s vice president of legal operations, agrees. “This will create a more data-driven discussion about what things cost,” Van Nice said.
Faster Review - For Nearly Everything
Generative AI can review NDAs for compliance against company standards. It can compare redlines against standard clauses. It can summarize key documents, so that human decision makers can act with unprecedented speed. For example, in the realm of contract management, once the model is trained on your documents, an in-house legal team could ask it to find certain terms, create redlines, and draft changes faster and more efficiently. BeiGene runs clinical trials across dozens of countries and many languages. “You need a small army to keep up and manage it in our CLM,” Ortman said. “I’m really hopeful generative AI will streamline the process of managing a lot of templates in our systems.”
Automating Legal Request Workflows
Generative AI models can build efficiencies into manual processes, such as tracking workflow data reporting. An attorney may take a day and a half to respond to an email request for advice. Generative AI can recognize what requests are coming in and automate the data entry and tracking. “This tool changes the paradigm,” said Eric Sedwick, senior director and head of legal operations at TIAA. “And it helps drive decisions along the way.”
Large language models and generative AI have applications across a broad stretch of legal work. Imagine if, for instance, in the course of litigation and investigations matters, legal teams could generate near-instant document summaries of key evidence, or query entire data sets of documents for insights using natural language. We’re on the cusp of a next-level transformation of the way lawyers do ediscovery.
What Legal Professionals Can Do Today — and Tomorrow
To be sure, we’re not at a point where you can put a request into ChatGPT and present its result as a final deliverable. There are privacy and security concerns that no legal department should overlook. The regulatory environment is still evolving. And the system is only as good as the data it’s trained on.
No one at CLOC advised that legal professionals rush to roll out AI or large language models. But the common thread was around the urgency to take this new tech seriously because a company that doesn’t have AI built in eventually will lose its competitive edge.
Education is the big thing right now. Understand the capabilities, recognize use cases, figure out where it’s applicable. Then educate others, especially your stakeholders, because, as one speaker noted, they’ll have unreasonable expectations and that fails a lot of projects.
Have a strategy. Set rules around what your company can and can't do with AI.
As always, start small, run a pilot, show proof of concept.
Test it out in a low-risk, controlled space. One legal ops leader targeted six months to consolidate 1,000 membership agreement templates, but won points with her GC when she was able to do it in three months with the help of an AI-powered tool.
Work with partners. When onboarding a new vendor, issue spot with them. Before signing a contract, ask the vendor for access and play with the tool. If you raise it as a burden of proof, you’ll be more confident in any recommendations you make to your GC.
Get your house in order. Focus on your AI program policies and procedures. And homogenize your data, so when ChatGPT gets better, you’re ready.
Read our CLOC takeaways post about generative AI’s potential to scale legal services and the impact on legal tech more broadly in “AI at CLOC 2023: The Dream and Reality.” And learn more about Everlaw's commitment to responsible AI principles here.