Getting Started with Generative AI in Your Department

As a technology and impact event on our industry, AI is somehow being both overhyped and underplayed at the same time. You have some who act as though it will transform every single aspect of our work, leaving our space unrecognizable. And you have others who suggest that it will be useful in some scenarios but will leave most of our work unchanged.

It is an exciting but confusing landscape, but one truth is undeniable: You have to act. This is not a situation where you can fall back on our usual legal operations playbook for new technologies, sitting on the sidelines and waiting for best practices to emerge.?

Corporate legal teams are already using AI-powered solutions to transform big parts of their workflow and process. They are achieving breakthrough results in core challenges, saving time and money and extending the reach and impact of their departments.

Most significantly of all, perhaps, is that these teams that have chosen to engage actively with AI have entered into a learning mode. They are gaining priceless comfort, experience, and insight with AI. They are on a learning curve already. The longer you wait to start yours, the harder it will be to change and grow.

Get moving and get focused!

So what is the bottom line for those of you who manage or help lead a legal team? What should you be doing right now?

Create a dedicated team.

Staying on top of the rapidly moving technology and evaluating new AI solutions requires real dedication and time commitment. You also want to avoid multiple team members going in various directions, all getting demos and trying out different tools without oversight. If you believe digital transformation is important to your organization and want to do it right, it will require making it a priority and putting the right level of time and energy behind it.

This is what legal operations does well, so if you don't already have a dedicated person in this role, now is a great time to establish it with a focus on AI, modernization, and transformation. And if you don't have the luxury of that headcount, remember that there are many external resources that you can now rely on to help with this work – consultants, managed services providers, flexible talent, and so on.

Find the problem first, think about AI second.

Don’t fall into the trap of chasing the technology rather than addressing your real problems. The current legal solution market is swamped with generative AI-enabled hammers searching for? nails. We’ve seen this before with plenty of other “shiny new thing” technologies, including the previous generation of AI… Many people rush into exploration and even implementation of trendy new applications without thinking through what they are actually trying to accomplish.

Focus on lower-risk and human assistance scenarios.

It’s important to keep in mind that generative AI is still new and has a tendency to hallucinate (e.g. return inaccurate and false results) at times. You also have to remember that in so much of what we do in legal, accuracy is critical… Overlooking a small detail, providing incorrect advice or research summaries, or choosing the wrong verbiage in a sensitive document could have pretty major consequences in terms of both cost and risk to your company. So, consider solutions that take a “human in the loop” approach where you can establish some checks and balances. Generative AI may not be ready to fly solo on these vital tasks, but it can make a great co-pilot, saving your people time and effort.

On the other hand, there might be low risk areas where you have tested and feel pretty good about the results, and these could be good places to unleash the AI to work on its own. We’ve seen a lot of success in organizations using our AI for 3rd party NDA review where the risk is lower and people are feeling comfortable with testing tools to manage that process given the low risk and the low job satisfaction related to doing that type of review.

Take a connected and integrated approach.

Generative AI is attractive because it can do things quickly that would normally be hard or time-consuming for a person. And that’s undeniably exciting… but if you look at the early scenarios of use that are emerging in legal, many of them are quite isolated. You might have someone pulling up a conversational agent to summarize a document, someone else getting help with a first draft, and so on. These interactions are useful but sharply limited. To unlock the real value of AI, you need to connect it to an end-to-end platform that connects to your workflows and existing business rules.

As an example, consider your contract work. Do you want your people using generic AI plugins to evaluate and generate legally sensitive terms and conditions? Just think how much more impact you could generate with an end-to-end system that adds AI capability within a rigorous framework.?

Get your data house in order.

Just as AI is much more powerful when connected to your business rules and systems (see above), it is vastly more impactful when pulling from your data. Data is the multiplier that drives impact of generative AI.

Consider this. Do you want to use models that have been trained on publicly available data? Or yours? You have to think about the whole assembly line of data and the systems and how they can work together, and then put AI on top… not focus on a flashy separate AI. That also allows changes and learnings that are made to be fed back into your systems and built upon.

Validating, rearchitecting, and cleaning up your data estate may not sound sexy, but it’s where I predict most of the real “AI work” will go. AI is this incredibly powerful extractor of meaning, value, and synthesis, but if it’s pulling from low-quality and incomplete content you will never be happy with your results. Getting this right isn’t just part of the challenge, it’s likely to be most of it.

High work and high rewards

If this all sounds like a lot of work, it’s because it is! The early flush of excitement of AI may have led some of us to think it was this magical new technology that would solve our problems with little effort. And it is incredibly powerful and it will transform our industry and our world… but if you want it to move you forward, and not sweep you aside, you have to understand how to engage with it and embed it within your organization and your systems.

It will take a lot of focused effort, thoughtful design, and perseverance, but if you get it right the rewards will be worth it!