Gen AI and IT – Transcript

By Tammy Pinter, Todd Musgrove, and Ron Exler
October 3, 2023
Season 5, Episode 1

Todd Musgrove:          

IT’s role is really to bring the technical expertise that’s needed, understanding the possibilities and the limitations of AI from a technical perspective. But you do want the business overall to have a strong representation in leadership. IT’s role is really to inform that and make sure that the business leaders are well-educated.

Announcer:

Welcome to The Hackett Group’s “Business Excelleration Podcast.” Each episode shares our latest research, benchmarks and best practices on world-class performance. Week after week, you’ll hear from top experts on how to avoid obstacles, manage detours and celebrate milestones on the journey to world-class performance.

Gary Baker:

On this week’s “Business Excelleration Podcast,” how can IT organizations most effectively vet, evaluate and integrate artificial intelligence into their enterprise infrastructure? This episode is part of our ongoing series on generative AI and its impact on business services. I am The Hackett Group Global Communications Director Gary Baker, and today you’ll be hearing from three of our IT leaders: Global IT Advisory Practice Leader Tammy Pinter, Senior Research Director Enterprise IT Ron Exler and Associate Principal Technology Transformation Todd Musgrove. Ron, the mic is yours.

Ron Exler:

Thanks, Gary. Todd and Tammy, great to talk to you today. What do you think IT executives are asking about relative to AI?

Todd Musgrove:

Well, Ron, I think IT executives are mostly concerned with how AI can create value for their organizations. They’re asking questions like: How can we leverage AI to optimize our operations? What are the potential risks, and what are the resources and skill sets needed? They’re also concerned with ethics and bias of AI and data privacy concerns, and the adaptability of these AI solutions to their current IT infrastructure.

Tammy Pinter:

Yeah, spot on. I also know that the C-levels are also concerned about ethics and biases like how do they manage that, and what do they need to put in place? Do they need to have an executive committee? Is it a committee that just falls in place? The executives are “Yes, I want to embrace this new technology, and I want to be the forefront of it,” but it’s also, “How do I implement that in my world?”

Todd Musgrove:          

Yeah, it’s a very dynamic, multidimensional situation with especially the latest topics in AI and making sure that organizations do have a committee that is cross-functional in nature, which I know we’ll talk about later. But that’s very important to make sure that all the considerations are covered.

Tammy Pinter:

Yep, I agree. I know that later in the podcast we talk a little bit about what is IT’s role, right? In all of AI, we have some differing opinions, right?

Ron Exler:

Yeah. Well, The Hackett Group also does research on a regular basis to tap into the pulse of what people are thinking. One of the research studies showed that 62% have started AI projects – and that means at least a pilot – but fewer than 10% have a fully developed enterprise AI strategy. So, Todd, starting with you, what role should IT departments take in developing an enterprise AI strategy?

Todd Musgrove:

IT departments should be at the heart of developing an enterprise AI strategy. They should be responsible for identifying the right AI tools, integrating them into existing systems, and ensuring data privacy and security, as well as addressing any technical issues that may arise. They also need to be involved in educating other departments about AI and facilitating the cross-functional departmental collaboration.

Tammy Pinter:

So this is where I differ a little bit from my colleagues. I agree with the structural foundation, but I believe that AI is going to be adopted a little bit differently throughout the organization. I think that departments are going to come to you at a separate time, meaning that it won’t be on a great timeline – it won’t be all at the same time. HR will come to you. Finance will come to you. Everybody will come to you. I don’t think actually that Todd differs too much from this. I think that we just need to be set up correctly of absorbing what the business wants on their timeline. Meaning that yes, in IT we set up the crux of how we do this, but we set it up based on when the business comes to us.

Todd Musgrove:

Yeah, that’s a good point. Just thinking about the stats behind our research, 62% have started AI projects, but only 10% have a fully developed enterprise AI strategy. I think it’s really each organization needs to understand where they want to be from an AI perspective. If they are pushing forward, having fragmented initiatives going on in the organization without an overall strategy is never a good idea. Being able to have that enterprise strategy in place and build from that it’s really important.

Tammy Pinter:

So we talked a little bit about committees – so what is your opinion about cross-functional committees? How should IT participate in those? Should they be a leader? Should they be a participant? Should they set them up? OK, I’m just going to say this once and for all. I think that committees should already exist about new and upcoming technology, but this has taken companies by storm. So maybe we need to take a step back. But what is your opinion – by the way –Todd, of how do we deal with this right now? This is fast and furious.

Todd Musgrove:

I would say first of all if an organization already has a technology committee in general, you definitely want to leverage that committee, the maturity of the processes and governance behind that committee. There needs to be that committee or that capability developed – not only for AI, but for technology in general.

Tammy Pinter:

So based on everything that we’ve heard, Todd, I’m going to ask this. I’m going to be a little feisty. What do you think IT’s role in educating the enterprise should be about like generative AI? Are we a leader? Are we consumers? Are we just enabling the business? What should we do?

Todd Musgrove:

So I think IT’s role is really to bring the technical expertise that’s needed to the committees, understanding the possibilities and the limitations of AI from a technical perspective. I think also there is an opportunity for co-leadership, but you do want the business overall to have a strong representation and leadership in the direction. But IT’s role is really to inform that and make sure that the business leaders are well-educated.

Ron Exler:

I would add that I think because of, like Todd said, IT should be the one that knows the most about the technologies. They need to advise on the risks and let people know what this technology can and can’t do. I also think IT plays a key role in facilitating pilots and demonstrating use cases so that the organization can see what the benefits and the shortfalls of these technologies. Additionally, there are things like setting data governance policies for sourcing, labeling, things like that around the data, which a lot of these tools don’t do by default, but are important for enterprises to communicate to whoever’s using their solutions – whether it be employees or customers. Lastly, I would say IT needs to be integral in developing guidelines for use of the AI technologies, especially generative AI, which is new and for the most part doesn’t have such guidelines.

Tammy Pinter:

Ron, that is spot on. Ron, I know that you have another question when it comes to Hackett AI research polls. So do you want to put that out there?

Ron Exler:

Yes. There was a recent data collection that we did on this topic, and we found that chatbots really topped the list, especially those using traditional robotic process automation capabilities. Those are fairly widely deployed in places like contact centers and customer service applications. But we’re also seeing that many of these groups are using generative AI from an experimental perspective to see what it can do, and how it can supplement those more traditional RPA capabilities. Other AI technologies most often considered that we found through the research are machine learning, as well as image and speech recognition. So just those speech and image recognition capabilities can be important in different departments that are further along in implementing AI such as R&D and I mentioned contact centers, but also cybersecurity, customer service and help desk.

Todd Musgrove:          

Tammy, what technologies are you seeing being implemented from an advisory perspective with our clients?

Tammy Pinter:

So, I’m just going to say, depending on where you fall in the generative AI or just AI space, I’ve been strongly and vocally direct about this – that service providers are going to be the first to adopt this. The software providers – they’re going to enhance their products with AI. Now, I will say with our core group, they’re just trying to figure out, “How can we adopt this to make our world better – be more efficient?” Those types of things. I keep telling them, “Just hold off. Let your solution providers implement this in the solution that you already bought – the license you already buy.”

I know this is a controversial stance because everybody’s like: “How do I job generative AI? How do I do this?” I’m telling our clients, and Ron can correct me if he wants to, he can correct me and me saying that: “Hold off. It’ll be a part of your solution soon enough.” Most people don’t even realize, by the way, that generative AI is already part of the solutions they already bought – part of the license. If you have Microsoft right now – if you have a Microsoft Office and you have Edge and Bing – you have ChatGPT.

Ron Exler:

Yeah, I think you’re right. I think that where we’re going to see the widest adoption is through these solutions that folks are already using, including in IT for things like IT service management, where we’re seeing the major vendors insert some capabilities or enhance some capabilities that have AI wrapped around them. So I agree with you.

Tammy Pinter:

So, Ron, to that point, I will tell you that I feel like IT is actually one of the slowest adopters of generative AI. I’m like, we’ve already had AI in our business since the 1970s, OK? We’ve had bots. We’ve had this. We have DevOps. DevOps being one of the first adopters of generative AI, like, “Go do what I spend too much wasting my time on doing.” But it’s not as transformative as it can be in customer service areas in going to do other things. That’s more transformative, I guess. In IT, “You can do this for me, but I still have to check it.” We’ve been employing automation in our life for years because we’re the first department to get cut when it comes to cost savings every single time. This isn’t new to us. How can IT help with some of the challenges introduced by generative AI? Are we the leader or are we the enabler?

Ron Exler:

Yeah, I think IT definitely needs to take a leadership role in that. IT can illustrate and evaluate technology solutions for things like intellectual property protection, system and data security. There’s, of course, with generative AI, there’s concerns around plagiarism, and there’s some tools available to assess that and discover that. There’s also challenges around sharing incorrect information. All these large language models are good at having conversations, but may actually give the wrong information in response to a question, and IT needs to be able to set up the appropriate checking mechanisms and inform folks as to what the limitations of these systems are. You already talked a little bit, Tammy, about potential bias. Again, IT can take a leadership position in explaining what that means, and why it’s important, and what can be done about it. Also, in using these AI tools, it’s important to disclose that they are in use.

We saw an example recently where in New York they’re requiring now that if a human resources department is using AI for AI interviews or scanning resumes, things like that, they have to disclose it upfront so that the applicant knows that AI is part of the mix of what they’re dealing with. So things along those lines. This is still a new area, lots evolving, but IT can certainly play a leadership role there. Something else I would add is the human needs to be involved in AI as well. So this concept of human in the loop to verify AI outputs is something that IT should encourage. Lastly, there’s a lot of legacy systems still in play where AI may or may not be friendly with in terms of interoperability and integration. So IT is going to have to figure out how to make that work and explain that to employees. So those are some of the areas, and maybe Todd, you have some additional examples.

Tammy Pinter:

Literally. I agree, even though I have some differences of opinion. But Todd, what do you think as a practitioner every day working with clients, what do you think?

Todd Musgrove:

So, I would say the majority of understanding challenges overall, it’s just not generative AI, right? There’s every technology being implemented potentially has challenges. So I think IT’s responsibility is really staying on top of the latest developments in technology and being that trusted advisor to be able to help the organization understand how they can mitigate any of these challenges and identify the disruptions proactively as they’re implementing these new technologies.

Tammy Pinter:

So Todd, I’m just going to put this out here, and I’m putting probably Ron and Todd on the spot, but you feel like it just needs to be an advisor and take a step back, not a leader in this? Look, I’m challenged with this every day. Part of me is split between, “I want to be a leader. You’re doing this wrong.” Part of me is like, “I can’t lead every business function.” So, I’m just going to say this plainly, “Do you believe that every business function – and I do, I will say this, I do believe this – every business function needs to implement generative AI based on their need, their business function, what they need to deliver to the client, and that IT should just enable that?” I’ll just put it right out there. I believe that.

Ron Exler:

I think it depends. Tammy, one of the things you and I have been working on is this whole concept of AI ethics and committees that enterprises might want to or should establish to address the issues around ethics in AI. So I think that’s an area where I don’t think IT just wants to sit back and let each function do its own thing. I think that’s one of those cross-functional responsibilities where IT really should be in a leadership position and have an integral role. So I think it depends on what part of this you’re talking about, but there are certainly some areas where I believe IT should be front and center and leading. Todd, maybe you have some other ideas on that.

Todd Musgrove:

Yeah, well, the maturity of where each function is within an organization could vary, so IT may need to step up and take more of a leadership role in some areas if functions such as HR or operations or finance doesn’t have a perspective. I think IT should help step up and fill in that gap. Otherwise, they can support in a variety of ways and really help the business define their own destiny. Because if the business doesn’t lead and isn’t part of the plan, then the overall adoption of any of these solutions would fail.

Tammy Pinter:

So, enablement, it’s about enablement. OK, well, thank you, guys. Todd, Ron, I really appreciate you being here today. I’m just going to hand this back over to Gary to end our podcast for today, but it was very interesting. Thank you so much, and I always appreciate your insights. Gary?

Gary Baker:

Thanks, Tammy, Ron and Todd for your insights. Listeners can find other episodes in our generative AI series in their podcast feed or on our website.

Announcer:

Thanks for listening. We’d love to know what you thought of this episode. You can write us an email at podcast@thehackettgroup.com. If you like this episode, please share it. You can find the audio, helpful resources and a transcript of each episode at podcast.thehackettgroup.com. You can also subscribe at Apple Podcasts or your favorite listening app, so you never miss an episode. The Hackett Group is the global leader in defining and enabling world-class performance. Learn how we can assist with your improvement journey at www.thehackettgroup.com.