Latest / Herbert Smith Freehills Kramer Podcasts / Legal Tech Deciphered EP6: When Clients Communicate Everywhere, How Do You Collect Anything
Transcript
- 0:00Stephanie Barrett: From your perspective, when you hear the phrase, oh it's fine, it's just all the communications are in scope, uh, where does your mind go?
- 0:07David Beck: Uh it panics.
- 0:08David Beck: Um that's the challenge.
- 0:10David Beck: Yeah, years ago all communications really did just mean email.
- 0:13David Beck: Uh now it could mean Teams chats, teams channels, WhatsApp, signal, LinkedIn messages, Instagram messages, social media, direct messaging, text messages, collaboration platforms, and whatever new applications someone downloaded on their phone last week.
- 0:37Stephanie Barrett: Welcome back to the podcast Legal Tech Deciphered Trust, Truth and Hallucinations.
- 0:42Stephanie Barrett: Over the last few episodes, we've spent a lot of time talking about AI, data governance, and how organizations can use technology to get better insights from their information.
- 0:52Stephanie Barrett: In the last episode, in particular, we discussed the challenges companies still face in making sure they keep only the data they need and organise it efficiently so they can get the best value from it.
- 1:03Stephanie Barrett: But today we wanted to talk about something that perhaps isn't getting quite as much attention.
- 1:07Stephanie Barrett: We're hearing a lot about using AI to analyse data.
- 1:11Stephanie Barrett: What we're hearing much less about is the forgotten challenge beneath all the Gen AI excitement.
- 1:16Stephanie Barrett: Organisations are focused on analysing data with AI, while the process of identifying and collecting that data in the first place is becoming more complex every year.
- 1:26Stephanie Barrett: Before we get into that, let's quickly introduce who's joining the conversation today.
- 1:30Stephanie Barrett: I'm Stephanie Barrett, Director of Legal Technology across the UK, US, and EMIA.
- 1:36Stephanie Barrett: I help bring together e-discovery, legal workflows, and emerging technology, bridging the gap between legal practice and digital innovation.
- 1:44David Beck: Thanks, Steph.
- 1:45David Beck: I'm David Beck, head of eDiscovery in London, covering UK and EMIA.
- 1:49David Beck: I advise on data scoping and management, disclosure strategy, governance, and the effective use of advanced review technologies, including technology assisted review and emerging generative AI across complex disputes and investigations.
- 2:03Danbee Kim: Danbee Hello, I'm Danbee Kim, head of digital legal delivery in the US.
- 2:08Danbee Kim: I work with the legal teams at our firm on the side of discovery, investigations, and review.
- 2:15David Beck: Thanks, Danbee.
- 2:16David Beck: So as you were saying, Steph, from an e-discovery perspective, the collection challenge isn't going away.
- 2:22David Beck: If anything, it's becoming more complicated every year.
- 2:26Danbee Kim: Indeed.
- 2:26Danbee Kim: Before you can use any advanced technology to analyze your data, you have to identify where it resides and you have to collect them and then make it reviewable, searchable, before you can apply any advanced Gen AI tools.
- 2:45Danbee Kim: That can mean confirming sources, collecting methods, date range format attachment, where is the attachment, whether the data can actually be used in the review.
- 2:56Danbee Kim: So that's uh you know, hardest part.
- 2:59Stephanie Barrett: Yeah, I think one of the things that's really changed is what we mean when we actually talk about something like communications.
- 3:05Stephanie Barrett: Um, the three of us have all been doing this for a very long time.
- 3:08Stephanie Barrett: And back in the day, I think the phrase, oh, we just need all the communications, that's what's in scope, was actually relatively straightforward.
- 3:15Stephanie Barrett: The options for how people communicated, but particularly in a work context, how people communicated was actually pretty limited.
- 3:23Stephanie Barrett: Um, David, Danbee, perhaps to kick off this discussion from your perspective, when you hear the phrase, oh it's fine, it's just all the communications are in scope, uh, where does your mind go?
- 3:35David Beck: It panics.
- 3:36David Beck: Um that's the challenge.
- 3:37David Beck: Yeah, years ago all communications really did just mean email.
- 3:40David Beck: Uh now it could mean Teams chats, Teams channels, WhatsApp signal, LinkedIn messages, Instagram messages, social media, direct messaging, text messages, collaboration platforms, and whatever new applications someone downloaded on their phone last week.
- 3:55David Beck: So the real issue is understanding what communication channels exist, which ones people are actually using.
- 4:01David Beck: And there's also the challenge of where those channels are being used and what is work-related and what's not.
- 4:08Danbee Kim: Yes.
- 4:08Danbee Kim: And then oftentimes when you actually start collecting with the custodians and interviewing them with what type of communication method that they're using, sometimes they themselves don't know what they are using.
- 4:23Danbee Kim: So it's the after we image the phone and then uh put it in the forensic reviewer and then looking at it, then they realize ah, I use this and that uh for the communication.
- 4:34Danbee Kim: So I think it is really uh important to know to verify where uh the communication is, what method the client or users use.
- 4:46Danbee Kim: And then at the same time, you have to know where the attachments and emoticons or images and other recordings or uh links that they are communicating together.
- 4:58Danbee Kim: That's another challenge that we have.
- 5:01Stephanie Barrett: If we think about the conversations that we have with clients at the start of a matter, so when we're doing the scoping exercise, I think how different it is now to what it would have been say even a few years ago.
- 5:12Stephanie Barrett: I think historically we would have focused much more on purely having a conversation with a client's IT team.
- 5:18Stephanie Barrett: There would be that sort of information governance um questionnaire that we may ask them to fill in, and we would be asking questions of the IT teams of, you know, how do you store your email data?
- 5:28Stephanie Barrett: How do we get access to it?
- 5:29Stephanie Barrett: But we've we'd already sort of have the answers of where those communications sat.
- 5:34Stephanie Barrett: It was just a question of you know minimal complexity over how they were stored and how best to access them.
- 5:40Stephanie Barrett: I think what's shifted now is that focus more on, to your point, Dan, is the individual conversations, those individual custodian interviews and questionnaires, where actually we are asking questions that are broader.
- 5:52Stephanie Barrett: So we're actually asking how do you communicate rather than where is your are your communications stored?
- 5:59David Beck: Yeah, that's right.
- 6:00David Beck: The number of potential locations just keeps growing.
- 6:04Stephanie Barrett: And in addition to all the different tools that people are using to communicate, um, we also have the additional fun and games of multiple devices and how individuals may be choosing to use those.
- 6:14Stephanie Barrett: Um obviously there's a variety of different options these days, but I think is it fair to say mobile devices still feel like the biggest contributor for us in terms of that complexity?
- 6:24David Beck: Yeah, absolutely.
- 6:25David Beck: The boundary between personal and business communication has become incredibly blurred.
- 6:29David Beck: People use the same phone for work, personal messaging, social media, and increasingly AI applications.
- 6:35David Beck: When you then factor also in that different countries have different privacy expectations, legal considerations on top of that, then when you add in cultural sensitivities, it becomes even more complex.
- 6:48Danbee Kim: And the technology itself makes it much harder for us to collect the data because it's keep advancing its uh privacy or security measures.
- 6:59Danbee Kim: Uh new iOS version has a much more uh difficulty uh getting through and get the uh reliable collection done.
- 7:07Danbee Kim: And encryption method is getting more advanced and harder to collect.
- 7:13Danbee Kim: For example, signal, usual method of just connecting and imaging doesn't work.
- 7:19Danbee Kim: You have to have a special device to go through or go beyond their encryption and collect the data.
- 7:27Danbee Kim: In an iCloud backup, you know, when there is an additional backup there, you know, how do you connect the messages and attachments and those relationships that you need to have for further analysis and review?
- 7:42Danbee Kim: That's getting more difficult as the time goes by.
- 7:46Danbee Kim: So it is really good for users.
- 7:49Danbee Kim: Personally, for me, I I love the security, I love privacy.
- 7:53Danbee Kim: However, as a as a practitioner here, it is uh more challenging and hard to collect the data reliably.
- 8:01Stephanie Barrett: There's a tension there, isn't there, in terms of the advancement in technology around um data security and encryption, as you say, Danbee, I think because so much more data is held by an individual on a particularly on a phone, um you want to make sure that that data is preserved, is encrypted, and you know, having things like stolen device protection on those phones is really important in terms of our day-to-day lives.
- 8:26Stephanie Barrett: But as you say, the challenge comes in when you are dealing with um legal processes, regulatory investigations, where large amounts of data may be in scope and you actually have to move very quickly to access that data, it can prove really challenging.
- 8:39Stephanie Barrett: And I always think it's it's interesting in terms of that conflict intention of we're heavily encouraging our clients to make sure that their data is safe.
- 8:46Stephanie Barrett: You know, if we think about the risks of cyber attacks, wanting to make sure the data is maintained securely.
- 8:51Stephanie Barrett: That's really important.
- 8:53Stephanie Barrett: But on the flip side, there's also then um a frustration when it is taking much longer than perhaps people would assume it would take to get our hands on the data and get it out of those environments because they've been made so secure.
- 9:06David Beck: Yeah, and of course that's that's even before you get to things like disappearing messages, different backup settings, or users configuring their devices in completely different ways.
- 9:16Stephanie Barrett: And obviously, we've talked a lot about the data.
- 9:18Stephanie Barrett: We haven't really tied back into obviously the context of all of these conversations, which is AI.
- 9:22Stephanie Barrett: Um, AI is also making some of these challenges even bigger.
- 9:26David Beck: Yeah, I think that's right, Steph.
- 9:28David Beck: We spent a lot of time talking about AI helping us analyse data, but AI is also creating new categories of potentially relevant information.
- 9:36David Beck: You've got prompts, the conversations, meeting transcripts, and AI-generated summaries.
- 9:42David Beck: All of these create additional information that organizations may need to consider.
- 9:46David Beck: And again, the data captured within those could be personal or work related, particularly when it comes to mobile devices.
- 9:53Danbee Kim: And we're also seeing meetings are routinely recorded and transcribed now.
- 9:57Danbee Kim: Uh often people want to generative AI output, transcript, summarize, next action items.
- 10:05Danbee Kim: Uh those were not um existing in even last year or a few years ago.
- 10:11Danbee Kim: We don't we don't have to consider that.
- 10:13Danbee Kim: But now we have to consider that whether it's a reliable data source that we need to collect, how to identify, how to maintain the relationship between different places where the transcript reside, where the recording reside, where the output reside, whether we need to collect prompts or tech conversation with your AI model or other AI-generated summaries and outputs.
- 10:41Danbee Kim: That that creates more complexity and more stories about it.
- 10:47Danbee Kim: However, whether it's a really right place or right information that needs to be collected for our uh discovery work, that's another human follow-up needed, uh, and then also more complexity added for our process.
- 11:02Stephanie Barrett: And we talk a lot about the technical advancement, but at its core, I think what we're we're saying is the impact that we're seeing is still human generated.
- 11:12Stephanie Barrett: It's it all of these changes, this creation of more data, more sources, more avenues for investigation and analysis, it's because of the change in human behaviour that AI has enabled.
- 11:24David Beck: Yeah, absolutely.
- 11:25David Beck: That data footprint just keeps getting bigger.
- 11:28David Beck: The collection challenge then also keeps getting harder.
- 11:31David Beck: And unfortunately, those two trends are happening at the same time.
- 11:34Stephanie Barrett: And one thing I worry about actually with this topic is um how it does get overshadowed by the AI conversation.
- 11:42Stephanie Barrett: Um, you see all these incredibly impressive headlines about how amazing, you know, and how fast AI analysis can be, how sophisticated the results are, how much more information that you can get out of your data sets now because of this technology advancement.
- 11:57Stephanie Barrett: And that may be true, but which that's all we tend to focus on.
- 12:00Stephanie Barrett: We tend to focus on the results.
- 12:02Stephanie Barrett: And when we um, you know, you see demos of the latest tools, you know, the latest functionalities that's available in e-discovery technology, for example.
- 12:10Stephanie Barrett: It's always showed um the phases are always showed after the data has been collected and after it's been processed.
- 12:16Stephanie Barrett: We don't always think about the fact that those advancements are also creating new challenges on the collection side.
- 12:24Stephanie Barrett: And we don't tend to see the focus on what happened before the AI was being run to show those really impressive results.
- 12:31David Beck: I think that's true.
- 12:32David Beck: When it comes to discovery and investigation, everyone wants to talk about AI review, AI investigations and productivity.
- 12:39David Beck: But before any of those tools can add value, someone still has to identify, preserve, and collect the data.
- 12:44David Beck: And that's still a fundamentally human process.
- 12:47David Beck: And as we discussed previously, AI can't help with data management, particularly well unless the underlying data is sound.
- 12:54David Beck: And anyway, that becomes less helpful when you're talking about data on individual user devices that have either limited or no corporate control.
- 13:03Danbee Kim: Yeah, I think that, you know, organizations and companies or you know, the people who are going through this uh the e-discovery process, they might assume that technology or the base of technology will solve this collection issue.
- 13:19Danbee Kim: Uh, but it's adding more complexity to it because the pace of changes or development of these communication platforms is extraordinary.
- 13:29Danbee Kim: It appears that you know it might we may cut up with those technologies, but the new uh applications and new technology keeps coming out.
- 13:40Danbee Kim: So this will add the challenge.
- 13:43Danbee Kim: But then even if you image and get the data, that's not the end of the process.
- 13:48Danbee Kim: You have to find a way to exporting uh and then connecting together with the uh the messages and attachments and other emoticons or any other reactions that come with all those chat messages, you need to find the way to collect and connect.
- 14:06Danbee Kim: So the more work starts from there as well to make the data um meaningful and then identifiable and then reviewable.
- 14:15David Beck: Yeah, that's that's right, Danbee.
- 14:17David Beck: It's a real challenge.
- 14:18David Beck: And the other thing I'd like to bring is something we mentioned earlier around cultural differences.
- 14:23David Beck: And one of those would always be a human process.
- 14:27David Beck: In that in the Middle East, for example, my experience was that a lot of the time where personal devices were involved, you'd have to segregate that personal and business information so that only the business information was put into the process.
- 14:39David Beck: And that has to be a human interaction, right?
- 14:42David Beck: It has to be that human element of actually making those decisions, usually sitting alongside the user of the device.
- 14:47David Beck: But that said, we shouldn't forget that the collection technology is working hard to keep up.
- 14:52David Beck: Relativity, for instance, can collect from various online communication and document repository portals, messaging platforms such as Slack, Bloomberg, and Teams, as well as data and metadata from Gemini and Claude.
- 15:05David Beck: Other tools such as Cellbright and Axiom are also increasing their capabilities when it comes to cloud and AI data source collections.
- 15:13Stephanie Barrett: There is perhaps the message that while AI is transforming what happens after collection, organizations still need to be thinking very carefully about the data that they're creating, where it sits, and how they will access it if they need to.
- 15:29David Beck: Exactly, Steph.
- 15:30David Beck: Good governance, clear policies, understanding what technology is being used, and being realistic about collection challenges remain just as important as they've ever been, possibly more important and certainly more complex.
- 15:42Stephanie Barrett: And maybe that's one of the takeaways from today's discussion.
- 15:45Stephanie Barrett: It's that the communication landscape continues to expand.
- 15:48Stephanie Barrett: People are communicating in more places, they're using more tools, and because of that, they're generating more information than ever before.
- 15:57Danbee Kim: Before organizations ask what AI can tell them about their data, they need to know where that data lives, whether they can collect it, whether it includes the attachment and related files, and whether it is actually ready to be analyzed.
- 16:14David Beck: Yeah, because when clients communicate everywhere, collecting anything becomes much harder than it first appears.
- 16:19David Beck: And applying the rigor that's needed to identifying those data sources and the approach of collection of that data is just even more critical than ever.
- 16:27Stephanie Barrett: And that's an excellent segue into our next episode, which is where we will go a little bit deeper into one specific area of this topic and explore how posts, private messages, ephemeral content, and platform-specific multimedia are beginning to appear more frequently in litigation and investigations.
- 16:45Stephanie Barrett: But we'll also look at why collecting, preserving, and authenticating that data is so much harder than it sounds.
- 16:50Stephanie Barrett: Um, with platforms evolving constantly and access routes often restricted or inconsistent.
- 16:56Stephanie Barrett: The theme of this conversation continues, and that is that AI cannot analyze what you cannot reliably collect.
- 17:03Stephanie Barrett: Thank you to everybody for listening.
- 17:05Stephanie Barrett: Thank you, David and Danbee, for joining me today, and we'll see you next time.
- 17:08David Beck: Thanks, Steph.
- 17:09David Beck: Bye, everybody.
- 17:10Stephanie Barrett: Bye.