OpenText’s Adam Kuhn talks about the growing relationships between law, IT and AI.

CCBJ: Tell us about some of the new trends that you’re seeing in e-discovery.

Adam Kuhn: Three trends I’m seeing in e-discovery include a bigger appetite for analytics and AI, simplified and consolidated procurement, and technology integrations.

Across the board, we’re seeing a growing appetite for much more sophisticated use of analytics and machine learning. We have clients that are using AI on every single project, either for prioritization, QC, culling or some other fact-finding or data reduction workflow. We’re having much more advanced conversations around workflows and capabilities than ever before, with clients being much bolder in pursuing innovative technology applications.

A new buying trend that we’re seeing at OpenText and responding to is the demand for a streamlined purchasing process with shorter contracts, a faster time to value and the ability to maintain flexibility with add-on capacity, services, etc. There’s an appetite for multi-matter packaging that strikes a balance between transactional and long-term subscription commitment, and we’ve recently rolled out a new package to address that.

Finally, there is a renewed emphasis on seamless enterprise discovery in regard to integration, both on premise and in the cloud. For most corporate e-discovery teams, their primary target data now lives in Office 365. This, in turn, is driving an integration strategy to connect e-discovery solutions with enterprise content management (ECM) solutions.

How are you seeing in-house law departments adjusting to the new technology and data sets that are available?

Corporate law departments are so much more sophisticated today, and they really have to be in light of the massive volumes, complex data formats and high pressure. Looking back just a few years ago, I think AI and advanced analytics were viewed with some healthy skepticism, either as a “We don’t need this” or “We can’t afford this.” But that’s definitely not the case anymore. The variety of data and the complexity of communications and language have really forced the issue on need. And the issue of cost has apparently taken a back seat, at least according to our 2018 Corporate Legal Operations survey that we conducted with Ari Kaplan. This year, only 63 percent of respondents indicated that they would use discovery analytics more if cost were not an issue (down from 92 percent in 2015). Respondents explained that more e-discovery tools are integrating analytics at no extra cost and also that the value is simply so plain – if not a necessity – that cost isn’t really a barrier.

What’s the role of AI in e-discovery and how has it evolved in recent years?

One of the most exciting parts of the e-discovery world is how the role of AI has grown and really enabled the legal profession to excel. Like I said earlier, there’s definitely an evolving approach and appetite when it comes to new technology. Back in 2012, the big question was, “Can I use artificial intelligence/predictive coding in e-discovery?” Now, in 2018, we’re asking more nuanced questions, like, “What’s the impact of a failed protocol on transparency obligations?” I think it’s a direct outgrowth of the enhanced sophistication of corporate clients.

The role of AI has changed as well. It’s evolved from the TAR 1.0 workflow that was very much a stabilization model. The TAR 2.0 workflows are generally more flexible and intuitive, and we see them used widely as a prioritization tool. Contemporary approaches to machine learning in e-discovery are able to learn continually from documents and decisions, constantly refining the model in step with the case team’s understanding.

We’re seeing clients use this technology on every matter because it’s now integrated and available to them at no additional cost. If for nothing else, they’re using it as a quality control check. And on top of that, I feel like the e-discovery “secret” is out – other departments and use cases are using these tools for due diligence, HR and privacy impact assessments, and internal investigations. As in-house law departments are able to insource a better technology like this – one that integrates all these tools – they’re able to add value across the board.

Let’s talk about AI in a different context. What do in-house law departments need to know about the different types of AI available to them?

AI is a big umbrella term. It encompasses several different technologies. Machine learning, for example, is a subset of AI and predictive coding or TAR is just a type of machine learning. That being said, there’re a few different flavors and they have different strengths and weaknesses.

At a high level, there’s supervised machine learning and unsupervised machine learning. The latter doesn’t rely on human feedback and can be used in-house to automatically categorize, organize and label large amounts of unstructured data into related concept groups, aka clusters. Not sure where to start an investigation? Concept grouping can be a helpful way to start slicing and profiling data beyond keywords.

The other form of AI, supervised machine learning, is the more traditional predictive coding or TAR approach, and it leverages a human feedback loop. A person looks at a document and says, “Yes, that’s relevant. No, that’s not relevant,” and the machine looks at those documents to build a data model. It’s not unlike Pandora or Netflix. I like James Bond movies and I like Jason Bourne movies, so Netflix might recommend “Mission Impossible” as the next movie to watch. We’re looking not just at individual words in a document, but phrases too, which is an important advantage when discerning meaning. For instance, the word “wind” in isolation could mean a lot of different things, but “wind power” is very different from “wind up” or even “wind down.”

How do you see IT and law departments collaborating on discovery and compliance initiatives?

In the past, IT and legal had somewhat of an adversarial or perhaps antagonistic relationship. Legal didn’t have the tools necessary to pull custodian laptops, collect the data and then process it, and so had to beg, borrow and steal IT resources wherever they could. In turn, those IT resources had to work even harder to complete their own projects. One of my favorite responses in our Corporate Legal Ops survey came from a director at a life sciences company saying: “We spend more time working with IT than anything on the planet, but it is the single worst experience of my life in terms of productivity.”

But that relationship is absolutely evolving now, with more understanding and collaboration between IT and law departments. Part of this is that many corporate legal operations groups have more of their own IT resources now. But equal is the shared understanding that designing and implementing integrated systems that enable self-service, seamless collections is a win-win.

With the growth of legal operations, we’re seeing more of a focus on data and metrics. What are some of the key metrics that legal operations professionals are focusing on and how are outside counsel responding to the new demands for these metrics?

We’ve been asking this question in our Corporate Legal Operations survey since 2015, and the first time we asked it the most tracked metric was data volume, followed by total e-discovery spend, and the least tracked metric was review efficiency. In our most recent legal ops survey, however, review efficiency tracking nearly doubled – 40 percent of our respondents are now tracking review efficiency. Interestingly enough, data volume dropped down in the rankings.

We also ask if corporate legal professionals feel comfortable with the level of e-discovery transparency and reporting provided by their outside counsel. In 2015, the results were dismal – only 28 percent felt that they had enough visibility. But this year there is a significant shift in sentiment: Forty-three percent now feel that they have enough insight into their outside counsel’s e-discovery processes.

I think this a really positive movement forward and also emblematic of the cultural change that corporate counsel are driving, not only in their own departments through the rise of corporate legal ops and process optimization, but with their outside counsel too through additional reporting and a heightened emphasis on technology use and efficiency.


About the author

Adam Kuhn - OpenText

Adam Kuhn is a senior product marketing manager and e-discovery attorney at OpenText Discovery. He holds an advanced certification for OpenText Axcelerate and is responsible for legal research, education. He frequently speaks at industry conferences on legal tech topics like machine learning and analytics. Reach him at

Leave a Comment