Artificial intelligence is far from a magic bullet that will solve all of corporate legal’s challenges or take over most of their responsibilities. But for many companies, AI can play a key role in enabling them to provide more cost-effective services. And perhaps nowhere is this more possible currently than in the area of contract management.
But while there are legal departments who have streamlined their entire contracting work through AI, many departments are taking a slower approach, cognizant of the challenges inherent in larger AI implementations. Speaking at various sessions at Corporate Legal Operations Consortium’s (CLOC) recent annual institute in Las Vegas, legal operations professionals from Microsoft and Cisco shed light on their AI contract management pilots and why, right now, they’re just testing the waters.
Tami Baddeley, operations lead, legal operations and contracting at Microsoft, for example, noted that her team is dipping its toes into contract AI technology with a six-month pilot project that leverages AI to analyze the contracts of its top 21 strategic partners. “We did a limited data set, so we are looking at 500 contracts. As we go through them, what we are looking for is, what are the terms that are common and what does our relationship look like based on these agreements?”
The goal of the pilot, she said, is to ensure that “in-house attorneys are no longer hunting for what their relationship looks like” and are instead focusing on higher-level work. The project will mainly be executed in conjunction with Pramata, Microsoft’s contract management vendor, which will analyze Microsoft’s contract data off-site using AI technology then send back reports to the company.
Deploying this relatively quick and small-scale pilot was pivotal for Microsoft as it looked to consider a larger AI contract management implementation because of the significant costs of such a project.
Baddeley noted that Microsoft wanted to first see how an AI contract management solution could work to benefit the legal department and work out its problems before committing any dedicated resources to a departmentwide deployment. Funding AI contract management solutions, after all, is not a one-time expense, but an “ongoing resourcing effort” that requires continuous funding, Baddeley said.
Mike Naughton, senior manager of legal shared solutions at Cisco, agreed: “This is not cheap; this is not something you will have to bang out in your spare time.” Like Microsoft, Cisco is developing a pilot to test how AI contract management solutions can streamline its contracting work. Cisco’s efforts are preliminary and are aimed at testing the waters before potentially approving a more widespread, and expensive, implementation.
But cost isn’t the only reason Cisco is taking it slow with contract AI. The company has also launched a second pilot to “get our hands on AI directly and see what it was capable of,” Naughton said.
The goal of this second pilot, he explained, was to “analyze individual contracts first without sending them to an intermediary vendor for processing or to an attorney to review. I wanted to see how AI [can] work on its own.”
As the first stage of this pilot, Cisco looked to apply AI to nondisclosure agreements (NDAs). Because NDAs are fairly short and simple contracts that are created and reviewed in high volumes, Naughton explained, using AI technology to streamline NDA processing can have obvious, quick benefits for any legal department.
Cisco’s legal team turned to Riverview Law, whom they are partnering with on their first pilot, to “set up an inbox where [the legal team] can drop an NDA in and within five minutes get an assessment of how the NDA fits within the standard,” Naughton said. He added that the assessment was directed by “what we call ‘a cheat sheet,’ which are really simple [instructions] around how we handle simple NDA terms.”
For Cisco, the second pilot was important in not only showing the legal team the abilities and limitations of AI contract solutions running independently, but also in allowing it to understand how to use AI in a more agile, targeted way before a more burdensome departmentwide implementation.
Naughton noted that the company’s pilot is an example other companies can follow to bring AI down to size and fit within their own workflows without having to first commit significant resources or organize all their contracts beforehand. Such a test project is important to those in the industry who “need to apply AI to the real world and don’t want to be doing some abstract AI [processes offsite],” he said.
Contact Rhys Dipshan at email@example.com. On Twitter: @R_Dipshan.