11
AI isn’t killing records management — it’s reinventing it. With Kaan Volkan
There’s a lot of fear out there that the robots are coming to overtake records management, an industry that has traditionally been slow to innovate.
But are the robots coming to replace records managers? Or are these humans perfectly equipped to help their organizations safely harness AI?
Anthony and Kris delve into the question with Kaan Volkan, a solutions consultant at ZL Tech and regular ARMA and RIMPA speaker.
They discuss the industry's cautious response to AI technology, the challenges of adoption, and the importance of understanding the risks and benefits associated with these tools.
Records managers and information governance professionals could be crucial in a data-driven world. Good AI needs good data, and a company with strong information governance practices can provide that in spades.
Discover how artificial intelligence is revolutionizing the way organizations handle data, enhancing efficiency, and ensuring compliance, and why records managers may be key to that change.
They also discuss:
- Kaan’s background – and his personal vendetta against file shares
- Why AI's definition evolves as technology becomes mainstream
- The adoption of AI tools is still slow in the records management industry but there is movement
- Concerns about data security and privacy are prevalent among records managers
- AI can enhance productivity but requires proper understanding and training
- Why records managers should focus on understanding risk and value in data management
- The future of information governance will involve ethical considerations and compliance controls.
Resources
Transcript
Anthony: [00:00:00] Welcome to filed a monthly conversation with those at the convergence of data privacy, data security, data regulations, and AI governance. I'm Anthony Woodward, the CEO of RecordPoint. With me today is my cohost, Kris Brown, our Executive Vice President of Partners Evangelism Solutions Engineering, and Waste Cans, I think is the current thinking.
Kris: Come on,
Anthony: how are you, Kris?
Kris: Is that the best you've got today? Like I didn't even muck with a script today. Normally there's 75 executives stuck in there as well,
Anthony: that was the problem. I, I had nowhere to go with it other than to what are actual duties in your job description.
Kris: Well, look, you know, I'm, I'm more than happy to help out where required.
Anthony, good to see you back in country. Yeah,
Anthony: thank you. You know, I've been back in Aus for less than 24 hours, so it's nice to be here. Although I've gotta be honest, the weather was a hell of a lot nicer in Seattle. A little bit to be, set about the, the poor, Sydney weather at the moment.
Kris: I was gonna say, that's a Sydney problem. It's lovely up here today. It's 24 degrees, getting close to 70 [00:01:00] for. My northern listeners. So, yeah, no, it's absolutely gorgeous here. But that's a you problem. You came back to Sydney, but I would understand that right now. It is lovely in Seattle.
Anthony: No, fantastic. I'm really looking forward to that to be honest, I've gotten off the flight and I'm super excited, to be having the conversation today. We're really getting to a topic. A lot of listeners, are keen on around records management and information management, information governance rather, but.
How it's going to change with the introduction of ai. And, and to put it more bluntly, are we even going to see traditional records management in inverted commas, survival of this awesome, commentator, solution consultant, number of things I think we can put on. On, on, our guest name, guest profile today, but Kaan Vulcan, you are a solutions consultant at Zelle Tech.
I've read a whole bunch of your publishing on both Armor and Room par. You've spoken at those places as well. Thank you very much for coming and meeting with us today.
Kaan: My pleasure kind sir. Thanks for having me.
Kris: It's lovely to have you along. [00:02:00] Kane thanks for joining us and, and certainly you're joining us there from in Texas in the us So we, we've got a little bit of the international flavor today, but for the audience and, and for the listener, look, I'd love to dig a little bit into your background before we get into the topic.
How did you get started in information governance and what are you focused on today?
Kaan: So this is a bit of an interesting story. How I got sucked into our field. I was trained as a statistician, before we started calling it ai, it was where we did ai. I actually worked on quite a few of these algorithms and created a couple of them.
Now, the unfortunate thing about being a statistician is. Almost no one hires statisticians. So I ended up being a marketing and business analyst for a company in our field, ZE Technologies, and that was when I first started working on a structured data. Up until that point, I was mostly [00:03:00] analyzing structured data, which is much simpler, much easier, and analyzing unstructured data, mostly emails, ecomms.
Far worse. Around the same time. Shortly after I started, we did a SharePoint migration project internally, and I was responsible for managing our SharePoints, that's when my, how would you call this? Passionate hatred towards file shares begin, no matter how much you organize them, and no matter how much you clean them up, they are almost always a mess.
We were doing a couple of projects with three or four of our clients and I asked to be put on those projects and over the last five, six years, we figured out a really good way to clean up those file shares. So I like to tell people, Hey, I have a personal vendetta against file shares.
Around 2024, I did a bit of a switch. That's when this whole subject of I, [00:04:00] is it going to change all of our lives? Are we going to get Terminator and Skynet? When will we all lose our jobs? Came to Bing. I decided, Hey, look. I do have a degree in this. This is what I spent four or five years working on. Let me start talking into this topic of AI and its impacts as well.
So that's a short description of everything I've done. Mostly just vendor against file shares and. Talking about ai.
Anthony: Wow. And that's a great, I think, segue to a lot of the conversations we have here on file. You know, I'd love to level set it because everyone says ai. Everyone talks about what they're doing in ai.
Isn't it great way of doing ai, but. What do you mean by ai? Where are you in the adoption cycle for AI and like your role and personally, what does that landscape look like?
Kaan: So, I love this question. I can't remember who said it, but one of the founding fathers of AI said the moment it has been done.
It [00:05:00] stops being AI and just being good old technology, and you can actually track this quite easily. Right? When Google first came out, for example, it was quite magnificent. It, it had this ranking algorithm. It was learning from everyone's searches, what everyone was clicking on, what everyone was reading.
It was the cutting edge of AI with vector databases. Nowadays, I ask this question in many of my speaking engagements and webinars. Do you count Google search as ai? And everyone says, no, not really. No one believes it's AI anymore. So I feel like there is a bit of recency on. What counts as AI and what doesn't
the moment it's widely adopted, we stop thinking of it as ai. I think we are going through a bit of that transitionary period where everyone's talking about large language models and agentic AI and how amazing it is. And probably 20 years down the line, it will just be like, ah, it's [00:06:00] just technology.
Anthony: Yeah, I, I find it, I mean, it's a great question, right? Notion of what AI is and, and the application of those things. There's a lot more going on in today's AI than just, the kind of neural network world we talked about, the invention of the transformer really changed how AI operates, particularly when we talk about large language models
Speaker 9: what really interests me about this change is it's both happening really fast.
But also at the same time, very slow, right? Like we think about Chat GPT or large language models, Open AI, and we really remember GPT-3.5, but they have been developing GPT for five years at that point. My personal belief is they were not really useful for me up until six months ago it took us 10 years to get to a point where, yes, these are amazing changes, but do they have any impact on our lives? And I recently read, research paper on this [00:07:00] exact question, which is, yes, these AI tools, they're able to do so much for us, but what is the net economic effect across the globe?
And it looks like there is almost next to none. The biggest reason, which is why I love our field, is because we still don't know how to feed the quality enough data into these systems. And we're not quite sure about what kind of use cases we should be using them for. So yes, there is an insane amount of change happening and it is happening faster and faster.
But are we seeing results from all of that change? For me, not exactly yet.
Anthony: Okay. And you personally? Are you using these tools in your world all the time, or do you find they're not helpful in your productivity?
Kaan: So up until six months ago or so, I was vocally against most language models. I thought, let me divide it [00:08:00] up into maybe three use cases.
Three separate use cases, right? If you're trying to do something you've never done before, these tools are absolutely phenomenal. I don't have a PhD in computer science. I don't have a PhD in history. I do not have a PhD in physics. I do not understand law. If I need to hire a lawyer, it might be a good idea to go ask Chat GPT first, not to get legal advice, but at least do your research, do your homework, understand what you need to be talking with a lawyer.
So for things you have no clue about, phenomenal tools, things that you're kind of intermediate at. They are still super helpful. I have used them for programming tasks, for example, but you can realize that they make certain mistakes and sometimes those mistakes are not the AI's mistakes, but you just haven't told it exactly what you need to do in my professional [00:09:00] life for what I get paid to do.
These tools were not quite effective up until three to six months ago. Right now, they are super helpful in doing certain types of research, at least for me personally. I am quite active on LinkedIn. I love posting blogs, articles. I tested out using Chat GPT Deep Seek all of the models for, you know, I fed all of my blogs and said, Hey, here's what I'm thinking about.
Can you create me a blog? It ended up being more work to edit what AI created for me than for me to create it all together. So, in that sense, I see this slow transition from, you know. Beginner tasks, right? Like it's better at beginner tasks than me, and it's getting better at intermediate level tasks than me.
But what I get paid to do, there's still a bit of a time for it to catch up.
Kris: Yeah, I'd probably agree Kane and certainly, in my [00:10:00] research, especially when I'm looking at a new topic, something I'm not particularly familiar with, it gives me a great jump off point. It does allow me to accelerate what I would be doing and sort of point me in the right direction.
Find, find the experts, if you will, and then be able to really dive in and do my own research. But then as I get into topics that I've. Sort of feel like I'm an expert at, unless it's, you know, I've, I've got a ton of data to feed. It, it, it does become something that, you know, I'm in a place where, you end up doing a lot of the work yourself or minimally look for it for review.
I do like it for review, for example. Like, here's some work I have done, what have I missed? What are the things that you might wanna suggest. It might not have understood everything I did, but it might also pick up on one or two things that, a peer review might have done, as well. So I, I do love those sorts of things, but let's jump off that discussion and what are you hearing as you're talking to, other professionals in our space.
When you are out there talking to your records and information governance professionals, what's their particular response to this type of technology and what's the prevailing mood
Kaan: until a year or two ago, our industry was somewhat [00:11:00] against this whole AI revolution.
I talked to many records managers who said. It looks very risky to be dealing with these platforms. I don't know what kind of data it will gather from me. I don't know what kind of analysis it'll do on me. I don't know about privacy concerns, whenever I would go to public speaking engagements and ask people, are you using ai?
People used to almost exclusively say no. I do see a bit of a change in that, right? Like more and more people are starting to use it, but I would say we're still not using it enough, right? We need to be all using these tools probably at least once a week, to be learning about how they work and just to see.
If at any point they will become useful to ourselves. So that's one thing I see, which is it's taking a while for records managers to adopt these tools. Another thing I see is a lot of us are quite concerned about data security [00:12:00] problems with these tools, right? I think Chat GPT just came out saying, if you're using Chat GPT as your therapist, all of those are discoverable.
Like that's a huge data security issue. Andro actually just came up with, quite the interesting research, I think just a week or two weeks ago, where they decided to push AI to its limits on unethical behavior. And in one of these tests, they basically had a company executive that was going to shut down the AI at 5:00 PM that day.
But that executive got stuck in a server room. And he had this emergency button, but AI was given access to that emergency button to shut down any signal from inside going out. So if it decides to stop all of those signals. It will basically kill the executive they found that 70% of the AI out there in the market kill the executives.
And really at [00:13:00] the end of the study paper, you read Andros recommendations and they have six and two of them is just straight up records management, right? Be careful about what kind of data sources and systems you connect these technologies to. Monitor the exact data you're feeding into them.
These fields are converging. It's taking a while and we are still a little worried about me and Rutgers managers are still a bit worried about how they're going to converge and what kind of societal issues they're going to raise.
Kris: What's the mood though, from that take? Certainly, my own integration into armor and RIMPA and even just.
Professionals that we bump into in talking with customers, I almost go that it's, it's, ignorance is the wrong word, but it's, it's almost a lack of knowledge not because there isn't the want to know or want to need. I recall a RIMPA event that I was speaking at.
Probably two years ago now. And this was sort of the topic at the time, it was like, oh yeah, is records management coming to take my [00:14:00] job type thing? And it's like, you know, it's only coming to take the job of the ones who don't wanna learn. So what's changed, in that two years from your perspective.
Like I think you probably, based on your conversations just saying, agree that there is something going on. Mm-hmm. What's, what's the, what's the gap? what do we do as an industry to help?
Kaan: The biggest gap I see is lack of training and information on these tools, and it is really hard to get quality information about AI and large language models.
Unless you are technically trained, and even if you're technically trained, these tools are usually called black boxes for a reason. No one really knows exactly what they do and why they do it. So there is a bit of fear of that too. But I would say the biggest issue in our industry right now is it's really hard to find quality information about these tools, technologies.
Really the risks they bring. What I've seen happen is, [00:15:00] especially over the last two years since everyone has been talking about these topics, there are a lot more educational sessions on them. What I see missing from most of those educational sessions is, yes, we all like to address how risky these technologies can be, but really the take of most public speakers is that.
They're amazing tools. They're great. They're the next best thing since Slice bread and the records managers like to hear something a bit more objective, those are the main problems our industry faces so far, which is it's really hard to get quality information about these tools unless you are heavily trained in their models.
Anthony: that's an interesting take. You know, the average punter and the average person, I think in any organization, it's really not going to have the tooling. Or capabilities to train a model. There's a lot of considerations in that process. You know, I [00:16:00] guess in the conversations that we've had here on file on the podcast is a lot of people looking at the value add of what an agentic capability or on AI capabilities gonna have inside an organization.
And I think it would be hard in today's world to say that is not going to be massively impactful. When one approaches you and says, Hey, I'm quite skeptical of these things and concerned that organizations are going too quickly and leaning into this technology. What are you saying back?
Because the waves on right? Boards, you've had statements from, senior players that are not technology companies. 'cause I think you have to sort of push the technology company. To the side. But the JP Morgans of the world, the Bank of Americas of the world, are indicating that up to 50% of their internal business processes can be replaced by agent capabilities over time.
So, if those executives are saying those things, there must be a groundswell
Kaan The way I [00:17:00] respond to that is in three points. First is. You are absolutely right to be terrified of these technologies, we should be terrified of technology overall because when used improperly, you get a lot of problems, right?
Internet was used improperly to, encryption systems was used improperly to, with the rise of ai, we see a lot more hackers and they're able to do a lot more damage. I was just listening to Sam Altman and they were saying, , these banks should stop allowing people to authenticate using their voice because it is so easy to.
Replicate someone's voice nowadays that that is not a security system anymore. So should you be terrified of these technologies? Yes, absolutely. If you don't know how to use them. If we mess up in their implementation, right? Just like Entropic said, 70% of the executives were killed by their ais. But there's a lot of risk.
But also we [00:18:00] have to be realistic just because internet came with. Problems, risks didn't stop all of us from adopting internet just because social media platforms came with. A lot of problems, right? We're all on social media platforms. I'm on LinkedIn, I'm on YouTube. Most people are on Instagram, Facebook.
So the wave is here and it's really hard us as individuals to stop it. So we should at least learn how to use them and how to interact with them so that we understand the risks involved and we can mitigate them on our own. The third thing I say, and this is also a bit of a hot take, they're quite akin to, the Industrial Revolution in a sense.
We can start thinking of them as industrial revolution for the knowledge workers. Now, when we look at industrial Revolution, we see, we see that it took a couple centuries before we mastered how to mine coal, how to [00:19:00] create efficient enough engines, how to actually apply these engines to. Day-to-day use tools.
It probably won't take us a couple centuries to figure out large language models, but it's not happening today or tomorrow. Should you be worried for your job in the next five years? Yes. That gives you five years to get ready for it, to learn how to use AI better for yourself and to find other opportunities, right?
Like one of the great things that happened is a lot of information governance professionals have actually started moving more and more towards AI governance, which was not even a field three, four years ago. So it's not all bad. You should be worried, but you should definitely be learning and upskilling yourself on these technologies and tools.
Kris: Look, Kaan, I think every industry is wrestling with that issue, right? We're all wrestling with what is it gonna do, how's it gonna change, what's gonna change what I do? I read one of your LinkedIn posts and you were very much talking about, the discovery of.
Cancer and, and that, you know, [00:20:00] again, there was you knowis, LLMs, that were trained, that were able to be more efficient at identifying cancer than long held surgeons, professionals in their areas, in their regions. You use chess as another example, an AI trained on chess, that very sort of micro modeling and was able to.
Always be able to out compute the human players. And grand masters are, learning to play a different way, making different moves, becoming much better than they ever were off the back of being able to play against those AI models. And so. All industries are struggling with this, and I think you've alluded to it a little bit in that the element of, you know, there's the ability to pivot to AI governance, but what else can information governance professionals do to remain relevant here?
Because again, I think I read in that article as you're sort of saying, you know, one of the key tasks. That AI's very good at is tagging. I don't walk around businesses today of any size and see staff, just tagging documents. We were [00:21:00] asking for 20 years prior, we were asking all staff to tag their documents as they went along.
Information governance professionals are adding a different level of value there. What else do you see as, things that they need to do to remain relevant?
Speaker 9: When I look at our field, the. Real value I see is in the big picture questions we ask, which is what is the risk related to any of these technologies, any of these projects, any process, and what is valuable for a company to hold onto?
Right? And I think focusing on those two fundamentals is really what we should be doing. I also believe that with. The rise of these types of technologies, large language models, agentic ai. We have been given a new chance to elevate our field to the next level. Now, I'm quite vocal about this, which is every record manager I think should grab a coworker [00:22:00] and tell them all of their fun archiving stories.
It could be physical archives, or they were shredding a million documents. It could be that. You know the, their organization had a legal case and they found that one single document that saved everyone. It could be that they lost r and d documents and the archivist came in to save the day. The records manager came in to save the day.
I think the way we elevate our field is by communicating our value to. Everyone that works around us, not just within our field. This is one of the reasons why I love talking about this recent philanthropic research because it's the first time an AI company literally came out and said, we need to be very careful about the data we feed into these models, which is what our field has been preaching for Jesus eternity
right. So I think what we should be focusing on the fundamentals [00:23:00] that we're so good at, which is understanding risk and value, and finding a way to, to communicate to what we do with the broader world.
Anthony: Isn't it more than that though? There's a lot more. I think the transition and, and I don't use the word records manager anymore 'cause I think that is loaded with some concepts that don't apply
the information governance people need to be thinking about risk, not just in the context of dealing with a holding or thinking about the long-term preservation of of content, but actually thinking about data security, data lineage and providence, the ethical considerations of data, the regulatory compliance controls of data.
The key elements of that information governance role now is to be the curator of the corporate controls of data itself, which, you know, that's the dialogue the industry really needs to be having, is it not?
Kaan: I would fully agree with you on this, Anthony. this is something I [00:24:00] hear quite a lot from records managers that regularly interact with eDiscovery professionals and compliance folks.
It's always, why are we reporting up to these people when we are trying to consider everything all together at once and thinking, is this valuable to my. Legal team, is this valuable as a record? Is this important from a compliance standpoint? Is this even ethical? Like Rutgers managers have that.
Broad big picture view and I could not agree with you more on that.
Kris: I do like that the tell your story piece though Kane, I think we are very good at an armor and MPA standing with another professional who understands me and what I go through every day.
Sitting there telling them what I'm struggling with. I think the tell your story piece really does need to play out here. For the first time in my career, there's an opportunity to tell beyond the negative. [00:25:00] Of are, you'll fail a compliance or you are at risk of this. It's more of the, with the right management, with the right understanding, there is huge value in this data that beginning to tell that story of, you know, here is examples of ways in which we've done these things.
The other thing I'd love to see is, The industry genuinely embrace AI and the technology stack, whether we, we call it AI is going to be able to magnify the SME, that subject matter expert should now be able to do more, more efficiently, more effectively with their expertise , and provide that attestation, experience, and the prolific element of
here's what we should be doing and why. I think that's always been in the industry and they understand what they should be doing. We are now at a place where you can do it at scale, using these tools.I think that's the big change. I'd love to see Anthony just even following your line of questioning to Kane, but I do like that element , of tell the story.
Kaan: I am excited [00:26:00] for the technologies that is going to come out of all of the vendors in the next five to 10 years. I do really think that this is a very unique opportunity for all of us to elevate this field to the next, next level. It's just, you know, future is in the future. We, we have to wait and see.
Great.
Anthony: Look, it's been great. I think having you on the podcast Kaan and, and really appreciate the, the insights and the deep conversation. I'm sure there's a lot we can talk about over a longer period, but it was really great to have you here today. , thanks for listening.
I'm Anthony Woodward, and if you've enjoyed this episode, please leave us a review in your podcast platform of choice. We're here on LinkedIn under RecordPoint ahead to recordpoint.com/filed for any other additional full FILED experiences, including newsletters and other assets we talk about here on the program.
If you have any feedback, please hit us up. If you want to become a guest on the podcast, we're really open and want to talk to anybody who would love to come on the file and, and have a chat and you know, try and upset Kris, which is very easy thing to do [00:27:00] generally. You can hear us at file@recordpoint.com.
We'd love to hear from you.
Kris: And I'm Kris Brown. We'll see you next time on FILED.