The 2024 AI Readiness Survey: Are You Prepared for the AI Wave?

Digital transformation has become one of the buzziest of buzzwords. This podcast is where we decode and deconstruct digital technologies, strategies, and best practices that are impacting the way companies deliver value to their customers. No buzzwords, nobs, Let's get to it. Hello everyone, and welcome back to the Digital Decode podcast, brought to you by Presidio and our partners over at Cisco.

It's been a minute since our last episode, and that is because we have been working on a couple of big projects, one of which includes the report we are going to cover on today's episode, and that's the AI Readiness Report. We polled over a thousand IT decision makers on their plans, goals and fears around AI and jen AI and today our guests Rob Kim Persidio CTO and Dan Lorman Persidio Field CISO are joining me to discuss the results of the survey. We give a few insights into what the data is telling us about AI in the industry.

So Rob, I'll hand it off to you if you just want to give us a rundown of what we were seeing in this report. Sure, thanks, Alec. Yeah, I know it's been a while since the last one. Certainly I've been on and I know Dan Lorman has written a great blog just talking about some of the analysis that we were able to get from the Benchmark report, So please be on the lookout for that, and we'll include the link as part of the podcast.

But in terms of the main themes, I think there's five that were very much highlighted during the entire report. The first thing is AI is an absolute imperative for every organization, regardless of your sector. The reality is that I think all of us in technology believe AI is a seminal technology and consequential. The thing closest to being this consequential is probably the transistor, and certainly we see a lot of analogies that are made to cloud and cryptocurrency and a lot of these other technologies, but I think that pales in comparison to what we know the potential outcomes that AI is going to be able to help us derive across every industry, sector and vertical.

In fact, you know, you have ninety six percent of IT leaders that believe it's going to be the main component in offering a competitive advantage, and you know more than a third thirty six percent, saying that it's essential for them to stay ahead against competition. I think the second piece that is going to be a resident theme here is around data being the foundation, right and the impact of data quality, how accessible it is in terms of the democratization of that data, and also the concerns around the ability to be able to make sure we have relevancy from the insights that we're getting as part of the I adoptions.

Right, So we have eighty six percent of IT leaders that you know, reported a lot of areas that they feel are because of data governance and difficulties in gaining actual insights out of the data that they're seeing. The third preparedness gap. I think that's a big one that despite all the enthusiasm certainly organizations in their rush to adopt AI, they don't feel they're adequately prepared. And we're seeing fifty percent of companies that are coming out and saying, hey, we feel like we're adopting this technology a little bit too soon.

And I know my colleague Dan certainly sees that in the public sector space, risk management is number four. It's crucial for as much as we talk about utilizing AI as a platform and tool for innovation. We have to do so with governance, cybersecurity, data privacy, regulatory compliance, the rise of what I know, Dan talks about a ton bring your own AI post significant threats and so you have to make sure that you're managing things much more proactively. And then finally, the need for education and expertise.

Dan, I'm sure you as we tour the country in the world and talking with organizations around AI, they're even their perception of what AI is is different. So the need to make sure that we're integrating in the ability to educate and provide AI literacy to our organizational IT leaders so that they can not only invest in it from an impactful perspective, but then also do so safely. Any comments on some key findings. Yeah, it's great to be back with you guys, and to ask thanks so much for hosting this.

You know it's Robbie. You do a great job at that last point. You know, as I travel around the country talking the different you and I have been a number of executive briefings in the public and private sector where we've educated leaders. But some of the key findings and insights I think are really important.

One widely recognized potential of AI. I mean, I think really across the board, the growing, the adoption, the numbers in there just keep going up every time I see a report. It seems like more and more organizations are getting involved and want to get involved and learn more about, you know what, what the opportunities are essential for saying ahead. Thirty six percent of it leaders call AI essential to staying ahead in today's market, so it really is kind of an imperative.

It's also a top investment area. Sixty three percent report AI as their company's biggest area of investment. I think that number, you know, this data just keeps going up every time we see it, so more and more than even the ones that aren't necessarily invested you know in the past, are getting invested, They're getting more and more invested. I think as we head into twenty twenty five, you're going to see that number even grow higher, pretty much get a ubiquitous across the board.

One more you know, varied adoption across industries, and I think you know, we we mentioned the fact that finance often leads a AI adoption, followed by healthcare, government in many cases lags. Behind, and I think we talked about that a little bit more detail. But I think there's a huge amount of opportunity I think specifically for the government areas, you know, to learn from our private sector partners and and see, you know who's kind of gone before them, how did they overcome some of these hurdles we're going to be talking about, and how do they really have effective AI programs?

And so I think it really opens up the door for more opportunities. Yeah, I mean you talked about again, So this concept that we certainly proclaim quite a bit when it comes to I innovation, but with governance, right, and so based on all of the fervor and adoption that we're seeing, there are a ton of concerns and hurdles, right, So again we mentioned data security and privacy. It's the top concern for more than a third of the IT leaders that we spoke to, especially in particular highly regulated industries.

High implementation costs. Right, let's let's go. So you know, there's not there's a reason why sixty seven percent are saying, hey, we're the most expensive thing we're investing in his AI because it's not cheap, right, And so there's a significant worry for many of these organizations, will we get the payback that we think in terms of return on investment or in the case of public sector, at least the adoption of the citizen ry around these new services. Right twenty one percent that are emphasizing the need for careful planning and budgeting premature adoption.

We've certainly seen a lot of news around unfortunately various different companies showing up on the front page of the Wall Street Journal for not good reasons because of adopting the technology too quickly without the governance in place. Fifteen percent admitting that adopting AI before being more prepared I would say not even fully, but at least more prepared, it leads to implementation struggles and then obviously technical challenges and cybersecurity. I think from a technical challenge perspective, the interesting point is, like a lot of newer technologies, you tend to hype it up way more than what's actually achievable.

And I think now that we've had a little bit more time with the technology, we're actually figuring out what it is that we can do, and then based on that start to maybe readjust our expectations on what AI can deliver. And rob I have That's an interesting point. You know that there's a ton of these companies that AI was this new shiny thing and it kind of companies wanted to adopt it super super early just to get ahead of the curb and make sure that they were up with the times.

But we talk about preparedness. Is there anything that you're seeing with certain organizations or that was seen in the report that were more prepared in their adoption and what were those kind of things that you saw that made a company more prepared to adopt AI alec. I think the biggest thing, more than anything else, is like it starts with step one, which is education, do you actually know what it is? Right?

And what's interesting in a lot of our talks that we have with organizations is especially with senior leaders, because you know, no senior leader with a lot of great experience, especially from a technology perspective, wants to be seen as not knowing what's going on and sometimes maybe a little more reticent in asking questions. But I think it's incredibly important that the literacy are around what is actually what the technologies are. The difference is between them, right, it's certainly generative AI very different from MLAI, traditional AI deep learning, which is very different from predictive analytics, also very different from automation, and yet in many cases we're seeing the implementation of all of these various different techniques and approaches getting integrated.

And so the first thing you need to do is just understand the difference. The next is we are starting to see a shift i would say in clients that are saying, hey, before we do anything in AI, we got to make sure our data governance is in place. But the reality is that you know, as humans, we're pretty messy with the way that we manage data. We're not very good at necessarily following the rules, so to speak.

And so as the data is getting more messy, right, not in databases, in emails and ims and texts that are all providing rich insight, right, especially if we use it appropriately. And then on top of that all the machine generated content that's happening because of generative AI, there's no way to fully provide one hundred percent data governance. And so we are seeing this notion that how can we actually utilize modern techniques, including JENAI itself, to provide areas of governance, but enough governance in place that we feel comfortable to start to roll out some of these pocs and specific AI initiatives.

Yeah, I think I like just jump in real quick on that one point. You know, the surveys showed eighty four percent of organizations that adopt the JENAI. Experience issues with data sources. So, as Rob mentioned, there's a couple of ways you can slice that.

But I think beyond the education, the awareness and we can talk about security and privacy discussions, I think the data a data question is one that comes. Up a lot. Yeah, you mentioned so I know, Dan you talk about bring your own AI quite a bit, right, and so the idea that you know, one of the ways to kind of safeguard this would be, you know, how can we deploy generative AI projects but do so privately and right? And so this idea that let's utilize what's available publicly for education, literacy, initial R and D maybe qualify some of the dev but as we start to think about deploying this into production, you know, is there an alternative or option for us to look at private AI deployment, especially as you start to look at things in scale, not based on the training, which we're less concerned about from a GENAI perspective, right, we're using curated models from other other providers and really much more focused around the inference, which is serving obviously token capacity for our users.

Good point, great points, and on. Kind of along those same lines, I guess when you're talking about, you know, the differences between AI that we are using generative AI that bring your own AI kind of thing, what is the is are there any ethical considerations surrounding the use of generative AI that you are concerned about? And how can organizations ensure that they are using it responsibly with their employees. I'm going to let Dan handle out.

I think that might be a dann question. I mean, I think they're absolutely are ethical considerations, and I think it really does depends upon you know, you know, the organization you're in, So it really does start with the policy and the guardrails that you have in your culture and your organization. Obviously government, if we want to make sure that the fair and it's not biased, the data is not biased, right, we need to be thinking about Rob's absolutely right, you know, when when you think about having your.

Own private AI environment. But when we're also heard I was just in Boston, I've heard this actually probably six or seven digital summits and cyber summits around the country over the last couple of months. It's fall that, you know, a lot of people say, you. Know, even if we aren't ready to move forward, the employees are bringing this stuff in any way, So it's kind of like to bring your own AI world of you know, and the reality of it is there is a lot of those models that people are pulling from are off of the Internet, and the data could be from all over the place, so it's not necessarily you know, just a view from you know, Massachusetts.

Or Michigan, or what state you're in or what company you're in. If your employees are pulling in from you know, the wider Internet and all the data sources, you're going to have issues around hallucinations. You're going to have issues. I don't want to scare people too much, but you know, there could be challenges, right and so putting in those guardrails, putting in processes and procedures, policies, and thinking about the ethical implementations implications of.

That data, you know, is important. But there are tools and maybe we have another session we can dive into some more details, you know, where how you can protect yourself and what provisions and guardrails you can put in place to help you in those areas. You know, Dan, you and I, I know talk about this quite often, especially as we start to see more and more AI policies and mandates come out from state agencies, federal organizations, and then obviously even abroad when you start to look at EU AI Act and things of that sort, and how they relate to consumer protection specifically around data, and how you're seeing that landscape change quite a bit to the point where now data is becoming a commodity, right, a commodity that you can now create another channel of revenue in providing licensing access to your data to either larger providers from a training perspective, or in other ways to help data enrichment for you know, private organizations as well as public agencies.

But the one thing about policies, you know, that I always continue to preach is that it's great to have the rules and like these guidelines, but the reality is if you don't have anything there to enforce it, It's just like telling my kids, Hey, you got to ask your mom before you know cookie, if nobody's there, garden the cookie cooky box, you know now you're talking about something that's that that that you're paying a penalty for after it happens, and I know for many organizations we don't want to be in that position.

And just a quick comment on data bias and things of that'sort. I agree with you. I think I think the interesting thing is with. Generative AI, this concept of hallucinations and you know, any sort of prediction being wrong has you know, certainly elevated the reason why a lot of orgs don't want to utilize this technology.

The first thing I would say is humans make a lot of mistakes as well. That's right. I don't know of anything that we do one hundred percent accurately, and so I would say that it's important when you're looking at developing these use cases to consider what your actual benchmark performance is for the service that you're delivering now, right, Because if you're currently delivering ninety two percent effectiveness, you know, if I can do ninety two percent utilizing automated techniques and modern digital services including AI, but at the same time increased productivity and all the goodness that comes out of that, well, isn't that okay?

In terms of data buy all data can have bias, especially if you have confounding factors. It's evident in the data itself. Even if the data is has is pure and untouched, there could still be biases. And I do think, especially when you think about the impact of public sector and the various different agencies that need to support efforts across all of our citizens, that's where we have to really consider the bias equation more than more than anything else.

Absolutely, And that's actually a great segue, Rob, because when we're looking at this report, which will also be linked in the show notes of the podcast, so you guys can all go check that out and check out those numbers, there is this kind of difference when we're looking at it between the public versus the private sector, and so it highlights those differences an AI adoption between those two. And can we just kind of dive in. I mean, Rob, maybe you more so in the public sector and Dan more on the private sector.

But what are these differences? Are there any specific challenges or opportunities that are unique to the private and public sectors? Yeah, I think from a private sector perspective, there are two main kind of issues and themes that come out. First, it's just fomo if you're missing out right, we have to do an AI use case, and the number of I have three briefings with clients today lined up, and in every single case it's because they've been told, Hey, you're in charge of innovation.

And we need AI use cases. So that is something that we're seeing, and I think that's I think it's a mistake to start with the technology and not the organizational challenge. Every use case has should be focused on. What are the challenges that we should be addressing to provide better productivity, faster time to market in the case of public sector, the ability to provide better efficient and optimize services for our citizenry.

These tend These have to be the areas and then getting very very specific, not trying to solve for everything, getting very specific on the use cases. Those are the wins. And I'd say that the next piece, at least in the private sector, has been ROI right, Hey, this stuff is expensive, whether you get it as a public based curated cloud service like we see with Copilot and open AI and Anthropic and others, or we decide to build our own internally and download frozen models like Meta and run them internally.

The reality is like they're trying to figure out how do you beyond employee productivity knowledge worker productivity, gain some better efficiencies. And while there are definitely places where you can quantify that better, once you get past those initials around things like customer experience and things that we see around you know, contact center and places like that, it does get to be a little bit harder. We are starting to see some of these ROI use cases, ROI justified use cases starting to emerge more, and they do definitely tend to be much more industry specific.

Yeah, I would just added, you know, in government, you know, we certainly have leaders, followers and language as well. And you know there's you know, in California, Massachusetts, a number of states are really you know, forging ahead. I think, you know, I'm mainly spoke speaking to the SLED market here, state local government, education, doing some some discussions with the federal side. And there's a lot obviously a lot changing right now within your Trump administration coming in, uh, in the federal side, I think there's going to be a lot more to come on that.

I think the challenges are. You know, a lot of times governments are known for I come from government, say government background, CTOCSO in state state of Michigan, and we love to like do enterprise wide policies in place, and it's more like the stop sign first and then you know, slowly we go to yellow and then we go to green, rather than kind of you know, it's kind of say no and then ask questions later. There's a lot of that going on still in government, and part of that could be good.

That's where the b YO D comes in a little bit. Bring your own devices, bring your own AI b yo AI, you know, using tool sets to enforce the policy, as Rob mentioned, you know, to stop people from maybe using what I tell people, the traditional CASM cloud access security brokers. You know, we wanted to control where people went with their data. Maybe we said you could go to certain you know, portals, but we only wanted you to go to ones that were authorized, that were secured, that we knew were end to end security in place.

We didn't want you to go into free China downloads, dot com, you know, so we were kind of blocking those things. And the same thing is now true of AI. There's a lot of tools out there, a lot of our partners with Presidio that offer tools. I'm not going to start naming vendors, but many of them that offer tool sets that help you kind of control be the traffic cop and say you can do this, but you can't do that.

I think there's a lot of that in government right now, I think on the leading end of this, so adopting. We're seeing more and more states, as Rob. Mentioned, going out to specific applications, whether that be in the healthcare area, whether that be in Medicare and medicaid, whether that be in criminal justice, within the different sectors, education sectors, and they're having more and more specific apps are coming out that are applied, you know, within government, in those different sectors within government, all the you know major ones.

Transportation is another big one right now, so you're seeing more and. More of that. I think it's twenty twenty five. You're going to see a lot of great case studies, a lot of great examples.

There's a lot of proof of concepts happening now. A lot of people are going beyond that and actually adopting things and rolling things out. So I really believe twenty twenty five will be a year where you're going to see a surge in adoption of AI projects across state and local governments. Yeah, we've been.

It's funny we're calling it AI, and everybody obviously knows artificial intelligence, but I think for us, it's really more actionable intelligence, achievable intelligence that we're really seeing E merge. Now, Dan, I'm just curious. I wanted to ask you. You know, obviously we're pretty lucky here in the US that we have so much innovation and in many ways we're leading the efforts around AI.

You have to be mindful of being in the lead because you can always give up your lead. And we are starting to see from a regulatory perspective, discussions around right government regulation that may not be necessarily the best approaches to how we govern AI, meaning governing and putting in restrictions limitations on the innovation side of AI VERSUS much more focused on the protection and safety side, especially when you think about things like fair use in data. And I asked this because obviously in other parts of the world, right, in other countries, we're seeing the opposite, right Japan and China and some other countries that are basically providing almost you know, companies with protection from lawsuits and liability around use of data and training because they're trying to gain some advantages in terms of catching up, so to speak, and and certainly in the case of China, we're starting to see that with some of these models that are coming out of Baydoo and others.

What are your thoughts around that. Yeah, that's a great question, man. I think especially with the new Trump administration coming in, there's a lot of headlines around this, and I think we have to do another podcast in a few months after you know, after that, uh, with with Elon Musk doing the those you know, the Department of Government Efficiency and uh, you know, there's a lot of talk around less regulation and rolling back regulation. So I mean my gut going into it is you're going to see less regulation from the Trump administration that we would have seen possibly by a Biden or heerits administration.

But again, I think time will tell on that and and that leadership from the federal government that you know, maybe looking for some national policies and a lot of people were looking for some national direction on some of these topics. You know, it may not be coming from Washington. I think that there's going to be more we'll see. I mean, I guess my gut, my gut feel going into this as we're you know, as we're as we're talking here and before still in twenty twenty four, and you know, rolling this out, you know, prior to the holidays, is that you know, the reality is is that you know, there's there's gonna be a lot of changes coming for the federal government but also for some of those oversight organizations that apply to the private sector, that apply to stan local governments as well.

And so I'd love to hear your thoughts on that. What do you think. I mean, it's a little bit of crystal ball. I know we do big prediction to be poor.

I think, well, hopefully Alex will do another one in January. We're talking about predictions for twenty twenty five. Maybe we can dive into this a little further. But you know, we're just coming off an election where I think a lot's going to change in government.

Yeah, it's certainly it is going to be a I agree with you that it'll be a big shift. I guess my only maybe not a prediction, but my only wish would be, because obviously the government is into granting wishes to their citizen ry, would be that regulations are required. Oh yeah, military measures are required for AI for sure, but concentrate them in protection of the consumer and don't do anything to limit innovation, because I don't want to see this lead we have And I do believe, you know, you hear, you've heard it for many years that data is now the new oil.

But I do believe that people, the countries that lead the AI revolution will be the ones to be the future continue to be the future superpowers and and certainly selfishly, I guess I hope we stay on top. So I agree, and I would just say just one of the big things on this point is, you know, I think a lot of people are some people love Elon Musk, some people hate him. But you know, the role of you know, thinking about autonomous vehicles, there's a whole other area. It's not even about AI, but you know a lot of people think.

Autonomous vehicles and autonomous everything is going to be coming more because he's there. That could be good, that could be bad. You know, we're gonna need regulations around those kinds of issues. That's just one other example, but there's hundreds of examples like that.

You know, clearly, there needs to be direction, there needs to be policy, there needs to be regulation, around these topics. The question is, like you said, how you do that and how that gets rolled out? I think that'll be a big time in twenty twenty five. Absolutely, And you know this, I've said this on previous podcasts that you guys have been a part of.

AI is still very much and it's infancy. Maybe not it's infancy anymore, but maybe it's getting It's like you know, toddler legs learning how to walk, and so there's so many different areas where it can be improved, where it can be regulated. But when it comes down to organizations at their core, like we've talked about, we have to be prescriptive with AI. Not everything needs you know, certain AI engines, certain things you know, like other one business may need something different than another.

But how can these organizations ensure that their AI initiatives are aligned with their overall business strategy. In the end. Yeah, I mean, I think we've talked about a lot of the themes already. The biggest thing is, I would say, be much more focused around your organizational challenge.

In fact, it's interesting the number of times we have conversations with clients where we tend to talk about all the capabilities that we have. But I know for myself and I'm I'm sure you would agree. The conversations we have with our clients that are the most meaningful are the ones where they share the things that are causing them the most pain and the most frustration and frankly, the most monotony. And if we can then apply the capabilities that we have, AI being one of them, to help solve those challenges, that's when you get the value.

So don't be focused on trying to figure out what your AI use case is. Just be much more acute around what challenges you have as an org, and should AI have a piece in solving that solution, which I absolutely believe will be the case, then it'll figure itself out. The second thing is we've been talking a lot around use case and deployment for particular verticals and industries, but let's not ignore the fact that AI has had tremendous impact on platforms the way that we manage compute, storage, networking security.

You can't do security operations without AI anymore, it's not possible, right, and so the large adoption of AI technologies that are being infused into platforms is quite significant, and I do believe that it's getting to a point where within you know, you know, if you want to have a prediction, I think in the next twelve months, do we start to see more and more organizations adopt fully autonomic and fully autonomous systems in the way that they manage the deployment of those compute, storage and networking resources in a secured manner.

And so let's you know, pay particular attention on obviously how you can drive value based use cases utilizing AI, but then how we can drive those things to scale by utilizing AI, and how we handle operations. Well awesome. Do you guys have any final thoughts? We are rounding out the episode now, we're coming up on thirty minutes, but is there anything you guys want to add for us set you free.

I'll just say one last thing and just you know, as Rob mentioned earlier, but I just want to emphasize, you know, dive in. You know, wherever you're at in the journey, you know, whatever you're doing, there's an opportunity to be more efficient with your daily tasks, you know, utilize things. Sure you need to have appropriate guardrails, but we got to get to yes as well. We got to do it securely.

We got to add privacy. We got to make sure we're doing this in wise ways. There are some great opportunities for us to help you. We're here, We're available.

Bounce stuff off of us, you know. I know, Rob and I do this all the time. I just had lunch with the CISO from a major university yesterday. I mean, this happens all the time.

Happy to talk about bounce stuff off of us. You know. We love to engage with you and help you think about the journey. All right.

The only thing I want to add eleg and I might get myself in trouble is a go Penn state. We are. We can leave it in. Let's just say the lunch I was out was with another big ten university.

There was, but yeah, we work with all of them, so it's all good. Well I couldn't disagree more. Rob, I'm so sorry about that, but you know what, to each their own. Thank you guys so much for being on the podcast.

Listeners. You can find Rob and Dan and myself on LinkedIn, so please connect with us there. Like I said, we're going to have Dan's blog in the show notes. We will also have an executive summary blog.

We put on our site in the show notes, and a link to that final report so that you can go ahead and read that there. Thank you to our sponsor over at Cisco, and thank you to you for tuning in to another episode of The Digital Dcode. Navigating the digital transformation of your business is no joke. Prosidio can help.

Prosidio is a leading IT solutions provider that helps clients simplify IT complexity and drive return on IT investment. By investing in the future of IT solutions, we stay at the forefront of technology trends, which ensures that our clients have access to a wide range of technology solutions. Let Presidio help you tap into the extraordinary potential of a digital future. Learn more at presidio dot com.

You've been listening to the Digital Dcode For more glimpses into the digital future. Subscribe to the show on your favorite podcast player, and if you like the show, give us a rating.