Lawyers, unlike software engineers, may not automatically understand the benefits that artificial intelligence can bring to their organizations, yet the expectations for corporate legal departments to utilize AI have never been higher. The chance to automate simple, yet manual, labor-intensive tasks, such as e-discovery or legal billing review, while enabling human beings to work on more business-critical matters is driving enormous interest. Corporate legal departments are intrigued by the potential for AI to improve productivity, decrease costs, and save time.
Given the hype, it’s no surprise AI was a focus for attendees at the recent CLOC 2018 Institute. One especially engaging panel discussion featured Mike Naughton from Cisco Systems, Jeff Marple from Liberty Mutual, and industry expert Dan Katz. They joined Spotify’s Jennifer McCarron, who moderated the panel, in exploring how AI is being used within the corporate legal space. Top takeaways from that discussion included:
Everyone has a different definition of AI
Google defines AI as “the theory and development of computer systems able to perform tasks that normally require human intelligence.” But what does that mean for the legal industry? The panelists had diverse opinions.
Mike described Cisco as “an acquisition machine” that “puts forth contracts every five minutes” -- highly repetitive work that lends itself well to AI. He also pointed out Cisco’s “laser-focus” on how revenue is booked and said that AI is useful in automating those processes and helping deals close.
Jennifer also shared a story about how once, in a previous company, her team was asked to analyze and review thousands of contracts. They ended up hiring first and second-year associates to perform the overly tedious task. In retrospect, that would have been a perfect candidate for AI.
The panelists agreed that applying AI to contracts is a natural start. By using AI on repetitive and process-driven activities, you are in a better position to gain insights into situations you may not otherwise have been able to see. In many cases, AI will be able to spot anomalies that human beings may not pick up on.
AI can be a tough sell
AI poses challenges that can be amplified within legal organizations where the main focus is on legal matters, not technology implementation.
Panelists remarked that even very simple change management is still a challenge. Teams may experience different types of pushback from others in the organization. Some people may not understand the value or how to use the technology. Others may question the investment and whether or not it’s worth it, and we may still be a ways off from being able to truly sell AI’s value to internal constituents. As a result, some companies have had difficulty justifying AI from a budget perspective. We at WK have found tremendous ROI for AI-supported bill review that anyone looking to justify costs should read.
There are also some instances where AI may simply not be a good fit for the organization. For example, AI might be too big of a leap for corporate legal departments that do not have their simple “block and tackle” strategies down. Before jumping wholeheartedly into the AI fray, these organizations should first determine if they have any standard work or basic processes they can test it on.
Panelists offered some suggestions for organizations ready to implement AI.
Collect information. Their first recommendation was to make sure all documents are collected and stored digitally in a single location. Given the mass of information stored in corporate legal departments, this can be an enormous challenge, but it is also extremely important. Corporate legal teams must make sure they have their data ducks in a row before proceeding.
Start small. Begin by creating a roadmap that includes some easy wins up front. Work with general counsel to identify a couple of places where you can easily test for success. For example, Dan suggested starting with NDAs and acquisition contracts -- repeatable tasks that can easily be automated and tested. And, as previously mentioned, legal bill review provides a great opportunity to evaluate AI, especially when a proof of concept (POC) is offered to validate the potential impact for your specific law department.
Clean your data. For data to be useful, it has to be clean. Organizations working with AI must first ensure that their information is as pure as possible. As Dan stated, “If you need a hazmat suit to go in and clean up your data, this will be a lot of work. If your data is pretty clean, the cost will be much lower. In the end, though, data cleansing will need to happen eventually.”
Understand the indirect cost of time. Panelists felt that, when it comes to AI implementations, time was a bigger factor than money. As Jeff said, “It’s like building a house when you’ve never done it before. It takes a long time at first.” Mike added, “It’s not necessarily ‘direct costs,’ it’s the indirect costs associated with the time involved for attorneys, scientists, and others” that is most concerning.
More time, same lawyers
Perhaps the most common AI myth is that it will take the place of human beings. But AI will actually augment the roles that lawyers play within their organizations. It will support, rather than supplant, their jobs.
Jeff emphatically addressed this point. He stated that he does not have enough attorneys doing the work that the insurance industry currently requires. Implementing AI will get individuals away from doing repetitive work, so they can focus more on important work that will improve the bottom line, such as assessing risk.
I made similar points in an article I wrote for Law.com. No machine will ever replace the skills and knowledge of a seasoned attorney. Instead, it is that knowledge that will inform the machine so that it will ultimately be able to make better automated decisions and continue to be a component that propels corporate legal departments to higher levels of value.
I know we will continue to see how AI evolves within the legal and legal operations industry. And I also believe we will soon begin to wonder how we ever made do without all the efficiencies and connections it will bring.