Beyond the First Pass: How AI Is Changing the Work of Disputes Lawyers

---

Document review has long been a time-intensive part of disputes work. As AI speeds up finding evidence, connecting facts, and analysing data, legal value is shifting from locating information to interpreting its meaning and determining strategy.

In this exclusive Q&A with Future Disputes UK, Tripp Hemphill, Global VP of Enterprise Markets at CS DISCO, discusses how AI is changing case development, the need to verify AI results, and why flexible workflows are key to deriving real value from technology.

He also considers how senior lawyers can help the next generation develop sound judgment in an environment where AI is increasingly part of everyday legal work.

Enjoy the interview below.

---

1. How do you keep your own judgment sharp when more of the groundwork is being done by technology before it reaches you?

I use multiple AI tools every single day in my work at DISCO, and while I have found that they increase my output capabilities and improve the quality of my work, the true value I bring is discernment. I have to look at the content the AI serves up and make decisions about what's right, what's wrong, and what needs to be modified. It's easy when something is clearly right or wrong, but the real work happens in the middle ground—when the output is 80% to 90% right and good enough to massage and work with.

I actually think this process has caused me to sharpen my own judgment. Imagine you're in a room with five brilliant people discussing a topic. Suddenly you have five different perspectives, and you have to figure out which ones you agree or disagree with, and which ones you only partially agree with. Evaluating AI outputs is a similar exercise. When I ask for a summarisation, I have to sit back and think, "Wow, that's pretty good, but how could I make it better?" I have to ensure that nothing incorrect or lacking proper framing is sneaking through. Context is worth a hundred IQ points, and it's my job to provide that contextual enhancement when an AI confidently presents a statement that lacks it.

Working across multiple tools also forces you to learn what each one is good at, much like you do with colleagues. For example, Claude might be better at analytical aspects and boiling things down, while Gemini is better at drafting in a specific tone. Evaluating these strengths and weaknesses is another way to sharpen critical thinking skills. For mid-career professionals, evaluating AI-assisted work from colleagues – and deciding when to tell them to start over because they relied too heavily on the technology without using their own judgment – is going to be a major part of our jobs in the coming years.

2. In your experience, where does technology sharpen the way a case is built, and where is it still no substitute for human judgment?

Cases are built on facts, and facts are developed based on evidence. When you can get to that evidence in a matter of minutes and prompts—rather than the days, weeks, or months it took with earlier generation search and retrieval technology—it has a profound impact on your ability to sharpen the way your case is built. I have seen situations where a relatively light investment of time using tools like Cecilia Q&A has completely shifted a party's posture in a case, leading counsel to advise the client to make different decisions about how to pursue the matter.

It used to be that the evidence that builds the facts was latent; it was out there, but you had to find it. Today, through technology, we can identify that evidence, connect the dots, find trends, and surface it to lawyers almost instantly. Lawyers are informed by an answer that takes them straight to the evidence, which they can then attribute to a fact used in case development. It’s a really exciting time to watch brilliant, evidence-hungry lawyers get the key documents they need with such incredible efficiency.

However, deciding what to do with those facts and that evidence—whether good, bad, or neutral—is still the core of lawyering. That’s the human judgment component. I think most lawyers went to law school not to hunt for evidence, but to take that evidence forward, organize it into facts, and have it inform their case strategy and discussions with their clients. We are not hooking up to the brain and searching it just yet, so that critical strategic application remains entirely human.

3. When a new tool enters your workflow, how do you decide whether to trust its output enough to rely on it in front of a tribunal or a court?

The heart of all of it is "trust but verify." DISCO has created a way to deliver results, but then directly connect the ability to go and verify those results in a tightly integrated, easy-to-access moment. We do this through citations and documents that are served up alongside the results, through our GenAI-powered document review solution Auto Review, and through the narratives that are generated explaining why a tag was or was not applied to a particular document. Even if the evidence surfaced isn't quite the track the lawyer was trying to go down, the fact that the evidence backs up the narrative we provided is trust-generating in and of itself.

When it comes to tribunals and courts, everything must be validated. For example, there's new practice guidance in the commercial courts here in the UK that requires a third-party validation tool for your citations if you are using AI in any way in filings. Different courts are putting different guardrails in place to avoid the time-consuming and embarrassing errors we've seen firms make.

Ultimately, you can never let your guard down. You can't just package up the AI's output and send it off. You still have to read everything and bring human discernment, knowledge, experience, and context to evaluating the quality of an output before it goes anywhere.

4. Much of the judgment in disputes is built on doing the 'unglamorous' early work. If technology absorbs that work, how do you think the next generation builds the same instinct?

It is up to the more senior members of a legal team to ensure that young lawyers are getting the hands-on experience they need, whether that is with AI or without it. A couple of years ago, I hosted a roundtable where one general counsel was putting AI in the hands of junior staff and telling them to go to town (within privacy and security constraints), while another general counsel prohibited junior staff from touching AI so they wouldn't compromise their learning. I imagine those approaches have met in the middle by now as we've gained a deeper appreciation for the true benefits of these tools.

Guiding the early career journey is up to senior members, but is that any different than it has always been? Absolutely not. This moment in time just requires some different ways of thinking and some different ways of creating working opportunities for young people. We can't escape AI, so we have to figure out how to create the next generation of brilliant lawyers with these tools in play.

5. In your experience, what separates a team that uses technology well in a dispute from one that simply has the tools?

It comes down to curiosity, a willingness to experiment, and a broader appetite for change. The case teams that use technology well have an appetite to change the way they did things the last time—or the last 50 times—and to use workflow tools and resourcing differently. Teams that cling to the same playbook they've used 100 times before are the ones that simply have the tools but aren't optimising the opportunities or the value they could bring to the case and the client experience.

While it requires some risk appetite to evolve and advance, I'd argue it's riskier to keep doing things the same way. It's a mentality and a mindset. Every day I see case teams with these tools literally at their fingertips, but they either aren't using the generative AI capabilities or they aren't optimising them.

Having the tools and knowing how to use them is essential, but these tools are actually quite easy to use once you get the hang of them, and they are forgiving. You can iterate and learn. There is no other way to master a generative AI tool than to use it, play with it, see the output, modify your inputs, and observe the impact. That appetite and willingness to do something different and evolve versus accepting the status quo is the true differentiator.

---

For Tripp, the advantage of AI isn't simply doing more work faster. It's giving lawyers faster access to the evidence they need, while creating more space for the judgement, context, and strategy that remain fundamentally human.

Tripp will deliver the keynote " Why Doc Review is Fundamentally Changing - And How You Can Use it To Your Advantage" at Future Disputes UK this September, exploring how GenAI is reshaping document review and how disputes professionals can prepare for a shift towards higher-value judgement and strategy. 

Register now to join the discussion at Future Disputes UK.