Agentic AI and the Changing Data Risk Playbook: XChange 2026 Recap

‍

‍

Host: Fahad Diwan, Director of Product Marketing, Exterro

Guest: Bryant Bell, Senior Product Marketing Manager - eDiscovery, Exterro

Guest: Robert Bond, Senior Product Marketing Manager - Digital Forensics, Exterro

‍

Exterro product marketing leaders Fahad Diwan, Bryant Bell and Robert Bond recap Exterro XChange 2026, Exterro's annual global user conference for eDiscovery, forensics, privacy and compliance professionals, held September 21–23 in Nashville.

Across hundreds of customer conversations, one theme kept surfacing: teams are working to connect eDiscovery, privacy, forensics and compliance as investigations and data requests increasingly cross departments. When evidence moves between tools and teams, its integrity has to hold, and how it was collected, analyzed and shared has to be documented. The discussion covers how legal and forensic teams handle litigation, regulatory requests, suspected breaches and employee misconduct: finding relevant evidence, establishing what happened, where AI speeds that up, and what a practitioner still has to review and verify.

They also discuss Exterro ARMOUR, Exterro's agentic AI technology for legal, forensic and governance work. Rather than answering questions like a chatbot, it plans and carries out multi-step tasks — pulling in the right data and people, and pausing for human sign-off before anything consequential. The conversation covers where that reduces manual work across investigations and data requests, and where it doesn't. Anonymized customer results include document review in a fraud investigation cut from two weeks to about four hours across 200,000 documents, and document review for an HR data subject request cut from 60-plus hours to about three. Across three early adopter customers, teams report roughly 90% efficiency gains on matters and investigations.

For general counsel, legal ops leaders, privacy officers, forensic investigators, and anyone owning data risk across one or more disciplines.

‍

Subscribe on Your Preferred Podcast Platform

Apple Podcasts | Spotify | YouTube

‍

Episode Transcript

‍

speaker-0 (00:00)

Hello everyone and welcome to another episode of Data Xposure, the podcast for data risk leaders. We're live at XChange in Nashville, 2026. Some people call it Nash Vegas. I'm here with these two handsome gentlemen, my partners in crime. From product marketing for data governance. Robert is product marketing for digital forensics, and Bryant is product marketing for eDiscovery

‍

speaker-1 (00:22)

Yeah.

‍

speaker-0 (00:23)

and AI, AI, AI, AI. So

‍

All of you have been, both of you have been, excuse me, in a number of different conversations with customers. Robert, starting with you, what themes have you seen recurringly?

‍

speaker-2 (00:35)

Yeah, so with the law enforcement folks and the internal investigation folks, they're really starting to grab on to agentic AI. So they've used generative AI, right? To do searches, assist, go through search terms. They've become comfortable with that, but they're a little uncomfortable with autonomous AI because that's controlling how they do investigations.

‍

And you're talking about people that have to document every single piece of evidence that they find, where it came from, who put it on the computer, how it got there, what's it mean in the case. So autonomous is a little bit different for them. The thing is, the benefits are so overwhelming. When you think of a CSAM case and there's 10,000 images, it would take them three or four days to go through those images. They can do that now in three or four hours.

‍

So the benefits are so overwhelming, they're really starting to grab onto it.

‍

speaker-0 (01:32)

That's fascinating. Thanks, Robert. What about you, Bryant? What common themes have you s picked up on?

‍

speaker-1 (01:36)

Well, the common themes that that I've picked on here is really, we have a a large amount of people here, practitioners of legal hold and and eDiscovery. And watching their eyes sort of light up how ARMOUR can eliminate a lot of the time that they usually take keen things in or trying to navigate through the different applications. To them, I've heard things like that's a game changer.

‍

This is going to save me so much time. And then when they see the stats that we've had from our our beta customers or early adopter programs, and they see that those that the time savings are huge from, two weeks down to half a day to get things done, to them, that's that's amazing.

‍

speaker-0 (02:18)

And so for the benefit of our listeners who aren't here and who may not know as much about ARMOUR, can you just give a quick overview? What is ARMOUR?

‍

speaker-1 (02:27)

So Armour is your AI command center for the Extero platform. And what it does is that it actually has a natural language interface, just like you're used to, you can say used to in Google asking a question, or now used to in going to a say a CLOD or a chat GPT and saying, I need help with, fixing my car, right? Yeah. But in our context,

‍

You can go into Armour and say, I need to, I need to f to issue these legal holes to these custodians. Can you help me do this? I need to set up the reminders. And Armour will will walk through that process, help you do it, but it's even more than that. It takes that intent and puts it into action. And is actually, in a lot of respects, just like having a junior associate at your side who's

‍

Really, really fast. but being able to just type in what you want to do today and have ARMOUR do it for you so you don't have to go through the different menus. And plus with the MCPs and connections to other data sources, you really can get a lot of work done just through a single interface.

‍

speaker-0 (03:40)

Right, so it's AI that actually does the work, not just generates a bunch of text.

‍

speaker-1 (03:45)

Exactly. It exactly. So it's doing the work alongside you, serving up that context and and information that you need to make the judgments that that are that are really important for your corporation.

‍

speaker-0 (03:56)

Absolutely. And so one theme that I've seen that I think ties with Armour is this idea of convergence. Right? I'm seeing people from different departments realizing that they can't operate in silos anymore. They have to work together, especially when it comes to data governance, right? Because it's an organization's data, not necessarily anyone's department's data. And I think Armour might be really well suited at at helping organizations break down silos and have and leverage AI agents across

‍

various different workflows that cross departments. Is that something that you've seen in eDiscovery, this idea of convergence and using AI to break down silos?

‍

speaker-1 (04:33)

Absolutely. you can go down to, we just had the session on on DSAR and subpoena manager, right? With ARMOUR on top of that, it's really integrating your compliance department in with your eDiscovery department, and it's actually making that communication better and more meaningful between those two departments. Because you're not just getting you're not just having one department throw an email over the transmit to the other. You're actually you actually now have a platform where you can truly collaborate to get things done.

‍

speaker-0 (05:00)

Amazing. And what about you, Robert? What

‍

speaker-2 (05:02)

So one of the interesting things that you said about ARMOUR is it has so much power. It also gives our clients more use cases. So I'll give an example. We launched three different products in forensics a month and a half ago in July, right? 8.3, we upgraded. We got ARMOUR for OP. I'll tell you a little bit about that in a second. We have ARMOUR for FTK. ARMOUR for FTK is an absolute game changer for our corporate customers. And the reason why is it helps us.

‍

handle their incident response problems. So for example, you were talking about natural language commands, natural language questions. You can reach out with natural language questions to our endpoints. So we've always had a great history of quality agenic our agents. Now they're a genetic so you can ask an endpoint which is on a system, is there malware on the system? Okay. Has there been data exfiltrated from the system?

‍

That's something nobody else in the market does today, right? So you can pull that data. The agents no longer pull. They're immediately on those systems. And then you can say, hey, where has this come from? Where has it gone? And how do I stop it? And so incident response can jump in immediately, find out how much damage has been done, and like you said, bring in legal. Legal's gonna ask, okay.

‍

Document everything, we're gonna need that for insurance companies, right? Security is gonna say, okay, let's make sure everything is stopped at this point. We don't want anything leaving. And privacy is gonna say, was there any sensitive data taken? Right? So that's that convergence that you were talking.

‍

speaker-0 (06:45)

Fascinating points, Robert. And you identified a number of different use cases and you as well, Bryant, that AI agents can do well. through my conversations with people here, people are very excited about ARMOUR and the ability to use AI that actually does the work. There were also some concerns, right? About well, what should an AI do? And how is Exterra being responsible? Did you encounter any concerns and

‍

speaker-2 (07:07)

So I would say law enforcement is the laggard when you think about crossing the chasm or something like that with a new explosive, just life-changing technology, right? They have always gone from point A to point B to point C, right? Where's the evidence? Where did it come from? How did get there? Now that an autonomous system or an autonomous technology.

‍

Is actually coordinating the workflow, they still need to know how it's doing it, why it's doing it, what kind of evidence it's surfacing, and where that evidence came from. Because when they go to court and they're testifying, it can't be we used autonomous AI to do this case, right? It has to be, how did it get there, right? How can you tie it to this suspect?

‍

speaker-0 (07:57)

And what about you, Bryant?

‍

speaker-1 (07:58)

Yeah, well, Robert brings up a really good point that setters around trust, And when you talk about trust, I mean, trust has been built into the Extero platform for a long time. you have, you have you have the audit trails, you have chain of custody. All of this is built into the platform, and all of that is leveraged by ARMOUR. So, and even more, you used to have to go down to a log to see it.

‍

Now you can just ask ARMOUR, right? And it'll produce that report for you. do you want that in a bar chart? Do you want it in a pie chart? Yeah, you can get it any way you want, right? So but there is still that, there is still that, big leap into trust with when it comes to AI. And I think there's a lot of aspects around ARMOUR, the fact that it's only using the data that you give it rights to.

‍

that whoever's using ARMOUR at the time they're restrict restricted by our back. So there's guidelines and guardrails on on your data and the usage of AI in ARMOUR. And those are really important things. And being able to explain that but then also show it, especially to attorneys, which are very similar to law enforcement, they they then start building trust into the system.

‍

speaker-0 (09:14)

Thank you, Bryant. Well, there you have it, folks. Live from XChange 2026 in Nashville, aka Nash Vegas. the key themes we've seen, as you've heard, conversions. People don't want to work in silos anymore. They're realizing that in an organization, they all have to work together, especially when it comes to data. And folks are excited about leveraging agents as long as they can do it responsibly and the AI agents are trustworthy. Until next time, please like, follow, and share the podcast.

‍