Examining E-Discovery in Competition Law

Nathan: Hi, I’m Nathan Kanter.

Teona: And I'm Teona Climie, and welcome to this episode of the Blakes Sounds Business podcast.

Nathan: Today, we'll be exploring the e-discovery process used in legal proceedings to shed some light on what that involves.

Teona: We'll also look at how technology plays a crucial role in e-discovery. As technology improves, it allows our legal teams to work more effectively and deliver results faster for our clients.

Nathan: Here to delve into this topic are Blakes Partners Julia Potter and Jonathan Bitran from our Competition, Antitrust & Foreign Investment group, and Anne Glover, a Litigation Partner and head of the Blakes inSource team. Together, we will be discussing their experiences with e-discovery, particularly in competition law, and the technologies that are making a difference.

[music]

Nathan: Julia, for anyone outside of the legal industry who might not know what e-discovery means, can you provide a summary? 

Julia: Sure. So, at its core, I would say that e-discovery includes identifying, collecting, reviewing, and then producing relevant records that are essentially stored in an electronic format, and those are for legal proceedings.

And when I say “records,” that's a very expansive term. It includes emails, documents in all formats like Word documents, spreadsheets, PowerPoints. It also includes internal company messages, so the ones that are sent through Teams or WebEx or Slack messages or any of those kinds of platforms. It also includes voicemails, text messages, calendar appointments, images. It's really anything that you can think of that leaves a record and can be collected. Then, depending on whether that record is relevant or privileged, it might be producible.

You know, e-discovery has changed quite significantly over time. It used to be a much more manual process of lawyers reviewing physical materials and banker boxes and going through paper documents, but now it's a much more technology-focused approach where materials are collected electronically through IT processes. They're uploaded and ingested into major review systems and then reviewed through a platform by a team of lawyers. Those lawyers are often working remotely with the assistance of technology, and then they're ultimately produced electronically. 

Nathan: In what situations would someone need e-discovery?

Julia: So, e-discovery is used for litigation proceedings generally, and that could be civil or criminal matters. And then in the competition law context, e-discovery is typically used to respond to massive document requests that are issued by the Canadian Competition Bureau. Those requests often come up in the context of complex merger review, so you would receive what's called a supplementary information request that requires a whole lot of document production for a transaction.

Importantly, e-discovery is likely going to have a greater role in the Competition Act context going forward. There were recent amendments to the Competition Act, just from this last December, that gave the Commissioner of Competition formal powers to compel records for market studies on industries that the Bureau is looking to study. The document requests that come with them themselves are also quite broad and can require pretty extensive document production by potentially a whole host of industry participants.

The other important item to note is that there are also proposed amendments to the Competition Act that will be coming into force in the near future that will expand private parties' ability to bring actions for breaches of the Competition Act as well. So, of course, along with any new kinds of private actions, there will also be more document production, just like with any other type of litigation proceeding.

Teona: Jonathan, what are the steps involved in the e-discovery process? 

Jonathan: So, at a high level, there's identifying custodians, collection, ingesting and hosting the documents that are collected on the review platform, reviewing those documents, and then producing them.

The first one, identifying custodians — and “custodians” is just a fancy way of saying people who have relevant information. So, for example, in a competition merger review, it's the business executives who are running the businesses in question. We typically conduct interviews with them to find out where they store their documents, as Julia was talking about. IT has to actually collect those documents, and oftentimes, it can be done remotely.

Then, once the documents are collected, they have to be ingested into the review platform. The documents are deduplicated and organized such that there's only one unique copy of each document.

There's also a coding form which has to be set up, which is basically almost like a little panel where the reviewers can tag what they're seeing in the documents: relevance, privilege, maybe some other things that they want to make notes about.

Then the review begins. First, you've got to train the people who are going to be reviewing the documents. Here, at Blakes, we’re lucky to have people at inSource, who are trained in doing these types of reviews.

And then we also use machine learning to facilitate that review. For clients who want to make sure that every single document is reviewed, there's methods that involve eyes on every single document that's produced. And there's other methods where you train the algorithm and rely more on the machine to spit out the documents that are relevant — of course, with checking for privilege.

It can also take a day or two to run the production itself and create a final file that can be exported to the government or to the opposing party if it's a litigation, and then the process is complete. 

Teona: Thanks, Jonathan. Like in other legal procedures, there are no doubt challenges in managing e-discoveries. Can you give us some examples?

Jonathan: Well, first off, I think — and this is not just true of e-discovery — the more contingency planning you do, the better. So, I think some of the pitfalls that we've seen happen frequently.

One is late identification of custodians — that if you lose a few days or a few weeks just figuring out who those people are, then that can really put you behind the eight ball.

Another one is how to deal with documents that are relevant and non-privileged but are harmful to your case. And so, it's also good to have processes to try and identify — we call them, sometimes, “hot docs,” that could be negative things or also positive things that you want to make a note of to use later on and have a plan for how you're going to deal with those documents. So, if there are negative things, how can you put them into the proper context to mitigate that, and if positive things, how to use those to your advantage as you go through the process.

Another interesting one that we've dealt with over the years is, for example, if you're producing documents to the Competition Bureau, they're required to be produced in either English or French. So, you also sometimes have to build in time for translation, which can be quite complex and expensive, also if you're dealing with languages that are not as widely spoken.

The last piece in terms of timing is in competition merger reviews. There's typically a refresh obligation after 90 days where if your production isn't completed and submitted within 90 days of getting the request, then it has to be more current than of the date that you receive the request, which kind of creates this chicken and egg problem, where you're constantly trying to refresh the documents that you have.

So, in the competition context, it's very important to get it done quickly, otherwise it becomes more burdensome. 

Nathan: Anne, as the Firm's resident expert in e-discovery technology, how has e-discovery evolved in recent years as technology has advanced, and where is it heading? 

Anne: As Julia said, the focus in recent years in e-discovery has been more and more on the technology and how it can help complete the e-discovery task. The technology that has advanced over the last couple of years has allowed us to actually deal with the volumes that we are seeing.

The technology of the day is called “continuous active learning,” which is where you review a document, and you tell the computer if it's relevant or not relevant. And then in real time, it will then go look for other documents that are relevant or not relevant. And there's different ways you can set up the technology, but it's — certainly, the goal of it is to continuously teach the computer as you're learning, thereby gaining efficiencies and saving money for your client.

We expect to see generative AI coming into this space shortly. It's already there in the United States, but in Canada, we expect some of the software programs to have it within the next six months. The way we understand that the generative AI will work with respect to e-discovery will be that you'll be able to give the computer instructions about what is relevant and not relevant, and then as you learn more, you'll have to kind of reteach the computer, kind of as you would teach people. And over time, the computer will be smart enough to figure out what's relevant and not relevant.

There's also some talk about where, you know, other things that generative AI could do, and that would be whether or not it could be used to generate automatic privilege logs, and then also whether or not it would be helpful to identify key documents. And the problem with key documents is you don't really ever know what really is key. So, open question whether or not generative AI will help on that respect.

So, I think in the next year, we're going to see a lot of changes with respect to technology and e-discovery.

[music]

Nathan: Thank you, Julia, Jonathan and Anne, for joining us today and sharing your insights into the world of e-discoveries.

Teona: Listeners, for more information on this topic and our podcast, please visit blakes.com.

Nathan: Until next time, take care.