How to navigate the AI landscape

In January, Agiloft Chief Product Officer Andy Wishart got together with Jonny Badrock, Chief Legal Engineer & SVP of strategic partner SYKE, to discuss: the state of AI in contracting, what are some practical use cases of AI that organizations are already taking advantage of today, and where they see this hot-button topic evolving over time.

A look at AI today

Artificial Intelligence is a hot topic everywhere you look and has only continued to gain traction and noise, especially over the last year or so. Contracts are so nuanced, and each one is different, so are we really in a comfortable place to fuel the processes around contracting for such an important and intricate document or process with AI technology? Jonny explains that while we’ve been talking about AI for years, the technology has only recently caught up to the use cases that would prove valuable in our day-to-day lives. 

Jonny and Andy discuss the discrepancies between the attitude towards AI, your organization could lie on the side of AI’s poor brand reputation – is it accurate, will it take my job, is it market ready? While divergently, Jonny explains, we can go to the opposite end of the spectrum and fall into the unfortunate habit of putting the tech before the problem. If we’re not careful, we can find ourselves using AI as a verb, “Can I AI this?” or “What can I AI?” For both audiences, though, Jonny reminds us that you’re probably already using AI more than you realize. For example, are you leveraging machine learning to identify key terms, clauses, or dates within your contracts – that’s foundational AI! It’s important to remember that whether you’re in the hype or hesitancy phase of AI adoption, while the concept isn’t new by any means, we’re in an exciting time where the use cases will expand, and daily application will continue to grow and evolve. As we’re facing new problems to solve or looking for operational efficiencies across the organization, we have to decide, can AI be a part of that overall solution tool kit? What’s important about that statement? AI is a tool in your tool kit, it’s not designed to replace the mechanic, especially in contracting. 

How prevalent is AI usage right now?

Agiloft conducted a survey last year that looked at the attitude and usage towards AI across the contracting space. Legal Ops seem to be the most enthusiastic with 44% of respondents say they’re currently using AI, which wasn’t surprising to Andy or Jonny as they believe it applies mostly to data analysis and other project work like key clause extraction. Or, again, it could be a case of AI already being leveraged in the background of many of your day-to-day processes behind the scenes. 

Who is planning on using AI in the future? This is where it gets interesting, they explore an over 40% increase among lawyers that are looking to implement some sort of AI solution in the next year. Jonny credits this jump to the increased awareness and spotlight around AI and the overall evolution of practical use cases that technology experts are producing for the legal space. Andy credits this jump to the overall expectation from Artificial Intelligence, users are done reading, they want to see this technology in action – however, utilizing AI while remaining cognizant of the “human element” should be the driver in your decision making. So, there’s official evidence that the hunger for AI is there and as Andy mentions, it’s now the responsibility for SaaS companies like Agiloft to deliver against that hunger and expectation. 

How can I start using AI?

Beyond the interest gauged in Agiloft’s AI survey, participants were asked exactly how they’re using or would like to use AI moving forward. For this discussion, Andy and Jonny singled out four specific practical AI use cases: searching contracts, importing data from legacy contracts, tagging third-party contracts, and redlining through negotiations. 

1. Searching contracts

While searching contracts seems rather basic, Jonny mentions, there’s a lot of work that happens in order to set up a proper data foundation. This AI solution seems as if it’s almost a trigger event, contracts being searched for force majeure clauses post COVID, ESG requirements, GDPR, etc. Beyond events like these that made immediate search more imperative, Jonny mentions that most clients he works with through SYKE live the, “I’m not even sure where my contracts are, let alone what’s in them” life, and it’s time to not only digitize those documents, but make them actionable.  

What can an AI-powered search do for you? It’s not only a benefit to have a more powerful natural search functionality while looking for important data within a record, but Andy explains that AI-powered search functionality like Agiloft’s ConvoAI can take this a step further by generating search alternatives that could provide better results, or creating a summarization of key points within documents that are more easily searchable. 

2. Importing data from legacy contracts

“This is all about building that really valuable, solid data foundation,” Jonny opened, “CLM is fantastic for new contracts going forward and the amount of rich data that we can get…but a challenge will always surround those legacy contracts.” The reality is, he continued, we all have finite budgets and finite time and while this project seems like a large lift it’s important to remember the value of the data and insights that can be added across the portfolio of the organization. Anyone on the search for technology to solve an operational issue has a vested interest in setting up that technology for lifetime success, adding legacy contracts to your CLM and then leveraging AI to analyze and organize those contracts provides that valuable data foundation. Transform that mountain of data into true, actionable insights. 

3. Tagging third-party contracts under review

A use case that SYKE sees often when discussing AI with their customers is this concept of tagging third-party contracts. Similar to both searching and importing your own documents, why not have AI analyze inbound agreements? As Andy explains, if a CLM is seeing a contract for the first time, an AI analysis can identify the parties, and the key data points and clauses, treating the contract as if it was being generated from a template. Like other use cases described throughout the webinar, this can be seen as an initial and foundational element of AI implementation to leverage AI through things like, performing risk analysis on incoming agreements, comparing clauses to standard language, or auto-redlining contracts. 

4. Generative AI Redlining

Andy goes into depth, including a demo, while reviewing this use case which leverages AI to align and revise clauses throughout the redlining process. After analyzing the contract (thanks to leveraging AI for third-party review) you now can easily see how exactly the third-party document differs from your standard and preferred language. Agiloft’s GenAI redlining capability then allows you to create thoughtful insertions and deletions that align the documents, as opposed to simply replacing the flagged clause. Additionally, if language is flagged but there isn’t a standard clause to align against, the GenAI functionality can create brand new language based on a user prompt, like, “make this clause mutual.” 

What do we see for the future?

“When you peer into the crystal ball, what do you see?” Andy asks. Jonny reiterates that there’s a long way to go before AI matures as a technology in legal. We need to cut through the hype, harness the energy that’s there but ensure you’re using AI for the right use case, like leveraging GenAI for redlining. Interestingly, Jonny mentioned that GenAI redlining is an exact use case that he and Andy discussed over ten years ago, but it hasn’t been until recently that the technology has become scalable and usable in day-to-day activity.  

In the next 12-18 months? Jonny thinks it’s time to take action, the excitement is there so let’s execute and make these well researched and executed use cases into a staple of our daily routine. 

What do Andy and Jonny NOT predict for the future? The elimination of the human element. While they do see further work to automate the contracting process, and that there are abilities to introduce AI to some degree at every step of the contracting process, the important thing to remember is AI is a tool in your toolkit, not the end all be all. It’s important to keep that human in the loop for not only accuracy and a final review, but to help train your AI to develop some of those guardrails in the first place.  

This event was the first in a series of practical AI discussions that Andy has over the next few months – be sure to continue checking in to see how AI can and will evolve, even in just a few short months!

Watch the webinar with SYKE here!