Ep. 13: Do You Need to Hire a Chief AI Officer?

Summary

President Biden’s executive order mandating that every federal agency hire a chief AI officer highlights the growing need for this “new kid” in the C-suite. But does every company really need one? What are the demands of the role, and how do you hire (or afford) one? David Mathison, who’s been tracking chief AI officers, says the government’s action will accelerate hires in the private sector – and keep the U.S. at the forefront of an increasingly competitive market for this kind of talent.

Host: Mike Curran, vp of global talent, Tanium
Guest: David Mathison, founder, CDO Club

Show notes

For more info on chief AI officers, check out these articles in Focal Point, Tanium’s award-winning online cyber news magazine. You can also get the latest data on the position (and news on the upcoming chief AI officer summit) at the CDO Club, listed below.

• Preventing AI Bias Starts at the Top – Just Ask These Female Chief AI Officers | Tanium
• As Federal Hiring Deadline Looms, Chief AI Officers Are in Demand. Here’s Where to Find Yours | Tanium
• The Chief AI Officer Is Here. It’s Time to Make Room in the C-Suite | Tanium
• Chief AI Officer Summit – The CDO Club’s second annual CAIO summit will be held October 2, 2024 (a day after its Chief Digital/Data/Analytics Officer Summit), in Washington, D.C. Speakers will include the U.S. Treasury Department’s deputy CAIO, Brian Peretti; senior advisor of digital talent at the DoD’s Chief Digital and Artificial Intelligence Office, Angela Cough; and chief AI officers from Avanade, Fox Rothschild, and the U.S. Department of Housing and Urban Development, among others.
• Chief AI Officer (CAIO) Talent MapThe first-of-its-kind report produced by the CDO Club covering key chief AI officer demographic data, compensation, career path, reporting structures, and open CAIO jobs. Due out in mid-July.

Transcript

The following interview has been edited for clarity.

We’re already seeing the chief AI officer pioneered in technology companies, financial services, banking, healthcare, and now of course, government is going to be a huge sector…. “If you’re not on the AI bus, you’re under it,” is what I like to say. Those companies that do invest in this sooner will have an exponential advantage later.

Mike Curran: So the C-suite is getting pretty crowded these days. Just within IT, first there was the CIO, the Chief Information Officer; and then came the CTO, the CDO, and the CISO. Well, now there’s a new kid on the block: the chief artificial intelligence officer.

Hi, I’m Mike Curran, vp of global talent here at Tanium. And today on Let’s Converge, we’re talking about chief AIOs. What is a chief AI officer? What do they do, and do you need one?

Well, if you’re in the federal government, the answer is a resounding yes. President Biden issued an executive order [in 2023] that all federal agencies need to hire a CAIO to oversee the deployment of AI within those agencies. So today to talk with us about this phenomenon is David Mathison, founder and CEO of the CDO Club, which is the world’s largest community of C-suite digital and data leaders.

Welcome, David. Great to have you today.

David Mathison: Great to be here. Thanks so much, Mike.

Curran: Great. So I guess we should start probably with that executive order from the White House. How will that affect what will happen in the private sector? Have you already seen that bleed into the private sector?

Mathison: We were definitely pleased to see the White House’s executive order on AI. I think it’s a great mandate. And by the way, it reminded me of a couple of things… I started the first-ever event for chief digital officers in 2012, so about 10 or 11 years ago. And we created the first community for chief digital officers and ran the first summit for CDOs at Thomson Reuters in New York. That was 2013. In 2014, we went to the UK and we did the same event at the BBC.

And interestingly enough, in 2014 there was a guy named Mike Bracken – I’m not sure if anybody remembers him, but he was the executive director of government digital service in the UK at the time. And at that time, he mandated that all directorships have to hire chief data and chief digital officers. And it really jump-started the hiring of CDOs in the UK.

Another example, the second, was when the U.S. government mandated the hiring of chief data officers in 2019. We all probably remember the financial crisis of 2018 and then the government wanted more oversight and accountability in data. So I think those two data points you can see – will this move the needle in the private sector is your question. And I think just seeing those two examples both in the UK and in the U.S. for chief digital and chief data officers, it really expanded the opportunity.

Curran: Oh, that’s exciting. Interesting. Well, let’s spend a few minutes on the terms if we could, David, with CDO, CISO, obviously being the security – and the unique goals that a chief AI officer is going to have and where it’s going to sit in the C-suite and how it will interact with these other chiefs, if you will. Sounds like a lot of chiefs in the room at this point.

Mathison: In terms of hierarchy in the C-suite, both chief AI officer and chief data officers will vary depending on the organization structure and priorities and their turning radius. In some organizations, we’re seeing CAIOs and CDOs reporting directly to the CEO or the chief technology officer [CTO].

Chief AI officers’ responsibilities, they’re responsible for the development, implementation, governance of AI initiatives within an organization, usually across the organization. And the primary goal is to really leverage AI technologies to drive innovation, improve operational efficiencies, and create strategic business value.

I would also say they’re really becoming chief productivity officers, right? The goal is really to increase productivity and increase efficiencies by using exponential technology.

For chief data officers, by contrast, they’re primarily concerned with management, governance, utilization of data assets across the organization. So the role could be ensuring data quality and integrity, setting up data governance frameworks, fostering data-driven culture, et cetera.

So ultimately … the goal for management is to just ensure that both the CAIOs and the CDOs really have a seat at the table, have the ear of the CEO and the board, and that they’re empowered through budgets, resources, and headcount to drive AI and data strategies that align with the organization’s overall vision and goals and strategic value.

Curran: Yeah, it could be a scenario where maybe the CDO and the CAIO are combined into one individual, dependent on the skill set of that individual.

Mathison: Yes, we’re seeing that hybrid title, “chief data and AI officer,” [he chuckles] so more complicated acronyms – CDAIO.

Curran: E-I-E-I-O.

Mathison: I think we’ll see a lot of that. And a lot of the data folks are grabbing onto those responsibilities, which is a relief, because we really need some of them to step forward in this.

Curran: Absolutely. So the CAIOs have been around, but obviously it’s kind of bubbled up and it’s in a bit of a fervor right now. But how has the role changed in the last one to two years that you’ve seen?

Mathison: That is a great question. And as of one to two years ago, as you are rightly pointing out, GenAI ChatGPT, DALL-E 2, Midjourney – when all that came out, everything changed as far as requirements go.

So, a little bit about my background: We produce events all over the world, and in 2020, during the pandemic everyone retrenched obviously for the quarantine, but we were lucky to have a sponsor like IBM and we worked. We basically created a whole series of webinars that we delivered for IBM. We worked with IBM’s client AI accelerator team, where they would make their own technology internally and then share their findings with the world, which was just really great. And Inderpal Bhandari was the global chief data officer at the time. And so what I did was, just by way of background, built a number of webinars for them. And, to answer your question, it was really all about chief AI officers – folks who’d been deploying AI in the enterprise with 10, 15 years of experience, in computer science, ML, DL, and really strong experience with change management. Because a lot of these implementations can result in layoffs.

So, we did a webinar on AI in 2020, and then we did another one in 2021, and I suggested, Hey, why don’t we invite all chief AI officers to this one? I got chief AI officers from the Army, from WPP in London, from Michelin in France, and from Elements Health here in the States, and it went really, really well. So I had to start tracking chief AI officers. I think I was the only one on the planet who started tracking them. So I’ll give you some real numbers, Mike.

In 2020 when I started searching for IBM, I’d go on LinkedIn, put chief AI officer in quotes and select people, and there were 250 results.

In January of this year, there were 780. And as of today, if you search right now, you’ll get almost 900 results. [Publisher’s note: As of August 28, 2024, the number was 1,200.]

Curran: I do have a couple of questions about that, so we’ll talk about that in a little bit. But let’s spend a couple minutes on your organization, the CDO Club. You had your first – must’ve been very exciting – your first CAIO summit in December. How did that go? What industries were represented and verticals that maybe were more prominent at that summit?

Mathison: Great question. Having done events like this for 12 years all over the world, the excitement, the sponsors, the speakers, the delegates, I haven’t seen this kind of excitement even with chief digital and the height of chief data, there hasn’t been this much excitement about a technology from all of the stakeholders, including the press. I mean, we were quoted in the Sunday Financial Times. I appreciate Tanium’s early coverage of this, which really laid the groundwork for some of the future articles that have been written.

By the way, when I said there were 900 chief AI officers on LinkedIn right now, half of them are garbage. By that I mean, 50% of the chief AI officers you’re finding on LinkedIn, 50% of them I don’t count… I’m only counting chief AI officers that have roles at major organizations throughout the world. The other 50% would be mom-and-pop shops, people putting up chief AI. Back in the day, I was a “CTO and co-founder” of my company. Now we’re finding “chief AI officer and co-founder” titles. And so we’ve seen a lot of startups putting the title in. And then, sadly, there’s always the people who put chief AI officer in their title on LinkedIn just to get recruiters or hiring managers to follow them.

So having said all that, of the 500 people that we’re tracking, we know the sectors. So the sectors that were represented [at our CAIO summit] in Boston were technology, financial services and banking, healthcare, obviously academia, given all these initiatives at MIT and Harvard and Northeastern as well. There were folks from the government sector, and of course, no surprise, human resources and executive search were prominent in the audience as well. If the focus of Boston was 80% corporate and 10% nonprofit/government, 10% academia, the upcoming D.C. events are very specific: It’s going to have a government federal agency focus, of maybe 60 to 70%, we’ll bring in 20-30% corporates and another 10% in startups and nonprofit. But the real goal is: How can we help the United States government get the best possible people in these critical roles?

Curran: Yeah, it makes a lot of sense. And I really appreciate those numbers on LinkedIn – it shows the exponential growth and also that a lot of those could be fake profiles or profiles that don’t have a lot of substance behind them. But you talk about a real substantive profile. What are you looking at as far as what it would cost to hire a really qualified CAIO? There’s the early-in-career individuals that are maybe in the lab working on that next GenAI model or the LLM models, but they’re not necessarily CAIO candidates because they wouldn’t necessarily be able to do that change management you spoke about. But if you do get a good CAIO, I mean, what are you looking at as far as [salary]?

Mathison: We’ve been fortunate. I’m very fortunate because we do executive search as well. Our most recent placement was a CDO at Dow Jones, and I really want to dominate in executive search for AI this year. So we’re very, very close to this. I’m really grateful that Ryan Bulkoski, who’s a recruiter over at Heidrick & Struggles – every year, he comes out with a compensation report for CDAOs, chief data and analytics officers, and now he’s included salary and comp for chief AI officers as well. And what he’s seeing is similar to what I’m seeing in the market.

So for him, he’s saying that a reported average total compensation – that would include equity grants and other stuff – in the United States, it’s $1.1 million all in, and in Europe for the same CAIO title, it’s almost $600,000. From what I’m seeing, that’s for seasoned chief AI officers with a track record of success that have been incumbents at major sectors like tech, financial services, healthcare, et cetera. Obviously for government, it’s more in the $250-350K range, no stock options obviously there. Which is why we are launching a fractional chief AI officer service, because at $1.1 million, to try to, we’re trying to level the playing field for small to medium enterprises and nonprofits. So we believe that the answer could be CDOs taking the role, but also providing more fractional services to help small businesses get the talent without the cost.

Curran: Well, give me a short synopsis on that fractional CAIO. It’s really interesting. I think a lot of folks are going to want to try that before they take on all the financial burden of a full-fledged CAIO. How would you deploy that and what’s the best way for companies to kind of engage you for that?
Mathison: In my opinion – again, this is just my opinion – but having been through this now a couple of times with companies, the traditional model for executive search for C-suite executive search for this title in particular, it’s completely broken.

There’s only a few hundred CAIOs on earth, so who are you going to pull from? It’s not like CDOs back in the day, where there were thousands and thousands of these people to pull from.

And then I think, as you rightly pointed out in your note to me, who’s going to write the job description? I mean, who’s going to interview the candidates? You almost need a chief AI officer to do the job description and interview the candidates. And who’s defining the strategy? So I think the answer is pretty simple. You’ve had fractional roles, you still have fractional roles, for CFOs and CMOs, for small or medium businesses who just don’t need a full-time CFO, right? Or maybe they need a CFO to get ’em through an IPO, and then they need a different CFO. I think the model here is the same.

Providing fractional chief AI officers, the benefit to companies is you’re going to get immediate action and results. No need to wait 12 months for your candidate to start. Hey, by the way, we can place somebody immediately at your organization who can deliver a strategy, and if after three weeks or a month you don’t like them, we’ve got a whole bench of people you can replace them with.

There’s no offboarding, there’s no embarrassment, there’s no mixed expectations. I think that’s the big problem right now – there’s a set of mixed expectations between the CEO and the hiring managers and the CAIO and what they can really deliver.

Curran: So David, let’s say you get a candidate and it looks pretty solid on paper and you’re going to vet this person for their AI expertise. What are some of the bullets you need to really get an understanding? How do you vet somebody for their AI expertise?

Mathison: I’m going to separate it into two questions, because one is the chief AI officer traditional role, with 15 to 20 years of experience, like what we’ve been familiar with. And then the enterprise GenAI role, which has some different requirements. So I would say the natural next step is to look at their previous work history, check on their successes as well as potential failures. Make sure that they’ve had a history of developing AI strategies and delivering on them as far as business value to the company, making sure they understand how to match business value with shiny new toys. We don’t want people to get lost in every new LLM or every new iteration. Same thing with identifying AI opportunities: Look in their past history to make sure they’ve successfully identified areas where AI can be leveraged within the organization to achieve goals.

Of course, they’ve got to have strong project management skills because they’re going to need to oversee potentially distributed development environments. So overseeing the planning and execution of AI projects, what have they done in the past, including of course, all the data skills, data collection, model development, deployment.

Another real strong [requirement]: What is their background in ethics, responsibility, trust? These are huge issues. You’ve probably seen, as I have, a lot of the titles in the beginning for chief AI officers, especially in the federal government and the military, they still have “chief AI ethics officer” built right into the title. So are they familiar with compliance and ensuring that AI initiatives adhere to any regulations and ethical guidelines?

And then there’s talent acquisition: Can they bring a team? They’re not going to be on their own, and do they have a success of landing in a company and bringing an entire team with them, or at least being someone who can attract the talented team. That includes data scientists, not just AI people, but machine learning, AI researchers.

And then I would say, are they curious? This is a role where you absolutely have to stay curious and continuing education is just a requirement. You mentioned it earlier, with GenAI, everybody had to redo their business strategies. You just have to accept the fact that this is a role for lifelong education.

Then, after you look at all the technical skills, what are the C-level skills that they present? Because they’ve got to be able to talk to the board, to all the C-level members, and then also to all the constituents in the organization to explain the benefits of AI and outline the risks.

The other thing I would say is about ROI – being able to understand, to calculate return on investment, because we just don’t know what initiatives are going to be successful and what the costs are going to be.

So I would say those are some good questions for the traditional chief AI officer. I would tweak those a little bit for the GenAI folks because they’re not going to have 15, 20 years of experience in infusing AI at every major decision point in an organization. They may be really, really strong at GenAI but have no experience in change management or in data AI literacy throughout the organization or delivering a service throughout an entire organization.

You really need somebody who’s got good personnel skills. I hope I haven’t beat that horse too much.

Curran: Yeah, it makes a lot of sense, especially the change-management aspect for it, because change is a’comin’, that’s for sure.

Mathison: Yes. For sure. It’s here.

Curran: It’s here. That’s right. So I’m going to ask you now, where do we go from here? How do we evolve the next one to two years? How do you see things evolving here in this field?

Mathison: We’re already seeing the chief AI officer pioneered in technology companies, financial services, banking, healthcare. And now, of course, government is going to be a huge sector, and it’s eventually going to make its way to other countries, sectors, and industries as the need evolves. But in my opinion, “If you’re not on the AI bus, you’re under it,” is what I would say. [He laughs.] And I think, as I mentioned before, those companies that do invest in this sooner will have exponential advantage later.

And then I’m really hoping that small and medium enterprises and nonprofits and companies with smaller budgets will be able to use fractional services so that we can kind of level the playing field a little bit.

Curran: Yeah. Thanks, David. Well, I really enjoyed the conversation. As we finish up here, anything that we’ve missed or anything people should take away that maybe we haven’t touched on just yet?

Mathison: Sure. I know I mentioned it, but we produced the first CAIO summit in Boston, the next one’s in D.C. And we’ve got a CDAO summit, our 11th year. We’re really excited about holding the next one in D.C. The timing, the location, could not be better. We’re getting a lot of great press from it. And again, I hope we can get our community more opportunities to serve, because we’re in a very competitive world – this is not a joke, China, Russia, they’re taking this very seriously.

The United States, I believe, is at the forefront of all this right now, especially with these moves by the Biden-Harris administration. You don’t see a Mike Bracken in the UK mandating CAIOs in government positions. I think it’s a really smart move, and I think it’s really going to jumpstart a lot for us here in the States.

Curran: Fantastic. Hey, David, thank you very much, I really enjoyed this. Let’s keep in touch as this evolves, and it’s going to be evolving quickly, as you know. So thanks again for your time today. I really enjoyed this.

Mathison: My pleasure, Mike. Any time. I hope to see you in D.C.

Curran: I’ve been talking to David Mathison, founder and CEO of the CDO Club.

If you’d like to learn more about chief AI officers, check out Focal Point, Tanium’s award-winning online cyber news magazine. We’ve got links to relevant articles in the show notes. Just visit tanium.com/p for publications.

To hear more conversations with today’s top business leaders and security experts, make sure to subscribe to Let’s Converge on your favorite podcast app. And you can help us get the word out – just give us a five-star rating.

Thanks for listening. We look forward to sharing more cyber insights on the next episode of Let’s Converge.