Using Generative AI to Power Innovation in the Legal Sector with Lydia Petrakis, Corporate Counsel and Digital Strategist at Microsoft
In this episode of On Record PR, Gina Rubel goes on record with Lydia Petrakis, Corporate Counsel and Digital Strategist at Microsoft, to discuss how generative AI is transforming the work processes of in-house legal teams and law firms.
Gina Rubel: Lydia, I’m so excited to welcome you to the show to discuss the impact of generative AI on the practice of law. After your presentation on generative AI for the College of Law Practice Management (COLPM), I was intrigued by what you had to say.
What are the practical use cases for generative AI for in-house legal teams?
There are so many different practical use cases, and I love that you’re using the word practical cause that’s what we need to start focusing on.
How can we use it in our own legal work and not just in the generic areas?
I started thinking, what is generative AI and what is it good at? It’s a branch of artificial intelligence that helps us use natural language for understanding and generation.
What is it good at? It’s good at text generation summarization and answering questions so it can help us in practical ways in our legal work, whether that be research, drafting, negotiation, compliance or regulatory work, or litigation work. It helps us interact with this technology and a new way with natural language. That is the tool of our trade as lawyers.
So, I thought, what are some practical use cases that we’re doing here at Microsoft that might be helpful for others? I have four to share.
- Copilot. One of my favorite things about it is that it helps me to start my work at “good” or “great.” I am more likely to get to outstanding work product instead of starting at “average” and iterating 20 times. I use it to help create my first drafts, whether it be in a Word document, an email, or others. I just spend my day-to-day accelerating my content generation.
- Microsoft Teams. We’re so busy, and we all know that feeling of “Oh, I’m gonna be on three meetings at one place” or you’re 10 minutes late because the last one ran long and you want to know what happened. So I use it a lot in the meeting to help me catch up on anything I’ve missed to summarize and include action items.
- Q&A experiences. Our immigration group has done a fantastic job over the years of documenting their knowledge. We can’t get AI transformation without building on that foundation. Because they did all of that work, we have Q&A experiences that we’ve used generative AI for. It’s being able to reduce the amount of time to find knowledge and draft answers to client questions and help them at scale.
- Another area we’ve already started exploring is in contracting. That’s probably a space very familiar to many of us; it’s supporting others through analysis of documents and negotiation. We have an AI tool that has reduced our review time of these contracts by 75% by putting in a lot of automations and reviews, and now we’re starting to explore how we bring in generative AI to even reduce that further and also lots of other areas we’ve been exploring and contracting.
- Another area deals with the regulatory work we all are experiencing – that tsunami of regulations. We have a couple of different pilots and experiments going on internally with our own framework as well as with one of our partners.
We’re looking at bringing in generative AI to help with all sorts of different items, whether it’s looking at how our position might have changed over time, helping people quickly find answers and our regulatory knowledge, or even drafting first drafts of a lot of the pieces of our own governance framework for us.
How do you prioritize those use cases, and how do those priorities align with Microsoft’s organizational goals?
We started our journey first and wanted to make sure that we were thinking about our strategy and our prioritization as well as involving the diverse perspectives across the organization so we’re not leaving anyone out. We started first with the crowd-sourcing activity to ask our people, “Where are you already using AI? Where could generative AI bring additional value or new use cases completely?” We actually got about 250 responses. So impressive for a survey.
Through that, we were able to identify our three key strategic priority areas and we call them advise, transact and comply.
Think of advise as knowledge management, transact as contracting, and comply as a regulatory focus. We picked those three because not only did they show up as the consistent themes across all of those areas, but we also looked at where we could drive scale and impact widely across the department. Those are three areas where it’s not going to just help one or two people; it’s going to help hundreds or thousands of individuals with that workflow.
Are there misconceptions about the implementation of gen AI in legal?
Yes, of course there are, like with anything new. We all learn. We are learning together. Some of us think one way or the other. I think there are three main misconceptions.
One is that generative AI is just this magic tool that is going to come in and fix everything, and we don’t have to do any underlying work or change any behavior. I think that is a misconception because one of the most important parts of this is the data. I alluded to this earlier as building on digital transformation work. We are firm believers that we have to have our data in order. We need process automation helping us continue to gather and understand our data because AI is built on data. Did I say data enough?
Gina Rubel: I don’t know. I’ll go back to the transcript and have Copilot tell me how many times you said “data.”
Lydia Petrakis: It will do that for you and it will be a large count.
That’s really #1. Our organization is CELA – corporate, external and legal affairs. We’re calling it our CELA Data Factory. We’re thinking about how we’re going to identify all of our diverse data and unify and adjust it to power our experiences. It’s not magic that will fix everything. It is powerful, but you have to do some work with it.
The second misconception is that somebody else is going to do all the work for me. There’s like a data science team or legal OPS team, and they’re just going to solve all my problems. While all of those individuals and teams will completely help you on your journey, we have to have the secret recipe. The magic recipe is bringing that legal subject matter expertise alongside all of the technical expertise, and that’s where we see the biggest return on our investments – running these experiments with our legal professionals ready to lean in with us and our technical teams.
The last one is cultural transformation. Change management adoption is everything. We can bring all the best technology, but if we don’t bring people along, we’re not going to be as successful. It’s important to think about your strategy and how are you going to make it a department-led initiative bottom up and top down – how are you going to recognize and incentivize the people you want and give them the requisite skilling to be successful in this new world of AI.
In what ways can a shift toward value-based billing incentivize the adoption of generative AI technologies in the legal industry?
Our partnership between in-house and law firms is always a key important part of a strategy and how we’re thinking of moving forward because it’s such a close partnership. We work together all the time.
If you think about the canonical law firm, it’s billable hour and it doesn’t provide a lot of incentive to bringing in innovation to the practice because it’s thinking about revenue-based measurements. When we start thinking about value-based billing, I think that it can help us together.
It can create some additional incentives for us to work and experiment together and for our law firm partners to continue to innovate so they can have that profit and revenue while still looking at bringing in innovation and doing great work. I think we can also take some of our learnings from when we just shifted at the beginning from hourly-based to value-based billing. It’s been a tool to be used as an incentive for that area of partnership together.
In itself, value-based billing isn’t going to solve all of our problems. Just going from billable hours to an AFA is not going to be the turning point. It is having our in-house teams being clear with our law firm partners about what we need and want and expect, giving them the ability with policies and access to innovate, and then having continuous conversations together because we’re not going to get it right, we’re going to fail. We’re going to experiment. We have to give each other room to learn and grow together, and I think some of the value-based billing can bring that opportunity together.
How can law firms and corporations partner in generative AI innovation?
The key thing is to experiment together, and both sides need to work on it. We can’t just tell law firms, “Do it,” and sit there. And they can’t just wait for us to instruct exactly what we want. We have to partner together to have an ongoing dialogue when you run into a roadblock or you try something and it doesn’t work.
You can switch and be agile and move together, and I think one of the most important things to be careful about and the really big challenge to overcome is that we have to make sure that our risk aversion doesn’t completely stifle innovation.
That is not to say that there aren’t security, privacy, and confidentiality risks that we need to make sure that we’re being mindful of. We’re being smart. We’re taking into consideration that we’re making mitigations for it, but they shouldn’t be a blocker where it’s just a no-go. We have to make sure we can innovate and do it responsibly. You might have heard that at Microsoft, we’re very big on responsible AI principles.
Gina Rubel: I have heard that, and it’s interesting because I think about the different size firms and how innovation and experimentation can be costly. I do hope that there’s room for all size firms to experiment and innovate, depending on the types of practices they have and the types of work they do with their corporate clients. I am seeing a lot of Am Law 100 firms diving into innovation with their corporate clients, but there is a huge investment. This is just my hope for those Am Law 200 and smaller firms – that they do find the ability to invest in this type of innovation because it is changing the world within which we work and live.
What one piece of advice do you have for law firms about generative AI?
My one piece of advice is to get started. The pace of innovation is only going to get faster, and you have to start somewhere or you’re going to get left behind.
How do you envision the role of generative AI evolving in the legal sector, and are there use cases or technologies that you believe will significantly impact the industry?
I think that this era of generative AI is going to be completely transformative across industries all over. This is a unique time because we’re seeing the legal industry being more focused on technology. There’s more interest and excitement.
Our practice is right for this technology and the interface of natural language, so it is completely transformative. I think we’re going to see continued evolution. We’re already seeing great capabilities on the market and they’re continuously evolving and growing at an accelerated pace. There are great capabilities in the areas in our strategy of advise and comply, and we’re seeing more in transact. I think there are lots of great opportunities to use capabilities and experiment.
I think those two areas are going to take a little bit more work, especially for in-house because of a lot of our data is sitting in inboxes and isn’t just pretty and ready to go. But when we start getting those, I think that is going to be some of the most transformative work when we can help our attorneys and legal professionals across the department and our clients get the information they need. We’ll be able to generate that content, at least getting that 80% at the beginning.
Our professionals who have the most constrained resources, their subject matter expertise, can spend their time in that area we all went to law school for instead of the repetitive tasks. We’ve been saying it for a while, and we’re feeling it even more – the amount of complexity, the velocity of the business, and the demands on our legal departments are growing, and our resources aren’t keeping up.
Gina Rubel: I’ve been in the legal industry over 30 years and I look back at when we implemented email as a new technology and so many people thought, “Oh, it’s not going to change anything.” And it’s not only changed everything, it’s changed expectations and productivity. It’s now become unproductive.
It’s fascinating to me, and I can’t help but ponder if we didn’t have COVID and the experience we all had with going virtual and having to work from home and all of those things, whether many law firms would be as quick on the uptake with gen AI as they have been because we had to embrace other technologies that so many had discounted. It is evolving faster than any of us ever expected. We are on a major road of transformation in the legal industry, and for those firms that don’t embrace it, they’re not going to survive.
Lydia Petrakis: I couldn’t agree more with you, and I still am always floored and impressed at how fast we’re moving. If we look at the acceleration curve from introduction to technology to where it is today, we’ve never seen anything like this.
Funny point to your mention about COVID and Teams – I was recently doing a big presentation at our department, gathering our summit on AI for CELA and all the work we’re going to be doing on generative AI. I was reflecting back to the last time we were able to meet fully in person as a whole department; it was right before COVID, and I was presenting on why legal should be using Teams. I was on a mission, and we were working on that change management struggle to get people to change how they worked in Teams. I think that’s a perfect anecdote.
Gina Rubel: I appreciate that validation because it just confirms my suspicion. It’s certainly going to be a hot topic as we go through this year in 2024 and beyond. I do look forward to hearing you speak at ALM Legalweek this year. I will be speaking there the day before you on generative AI as well.
Learn more about Microsoft CELA: https://www.microsoft.com/en-us/legal/diversity/our-work-in-cela
Learn more about Gina Rubel
Listen to more episodes of On Record PR