17 Jan 2025 · 56m · Michael Krigsman, Adriana Karaboutis · 20 quoted facts
Boards of directors should push their companies to adopt AI, not block it, while keeping risk under control. Adriana Karaboutis argues that boards must play both offense and defense. They set strategy and risk limits, management runs the work, and boards need enough technology understanding to ask good questions without getting lost in technical detail.
The main tension is between innovation and risk. Boards worry about privacy, ethics, regulation and leaked data, but Karaboutis says that saying no is not safe either, because staff will use outside tools anyway. Her answer is to ask what it would take to reach a business outcome, then measure results with ordinary tools such as return on investment.
Karaboutis uses a sports analogy: boards should push the company to use AI to meet its goals while also keeping it within legal, ethical and regulatory limits. Leaning only one way risks either privacy and compliance problems or a lockdown that stalls the business.
She says the worst outcome is employees copying company data into outside tools, as happened at Samsung. Her fix is to bring in a company-controlled version of the tools and tell the workforce clearly what is allowed.
Management handles day-to-day execution and tool choices. The board steps in when initiatives are large, costly or central to strategy, and it checks progress, risk and spending against the commitments made.
Karaboutis says boards use return on investment, return on equity and key performance indicators, plus softer outcomes such as employee satisfaction. Even a learning goal, like raising expert staff from 10% to 20%, counts if it was stated up front.
She says one token technology director is not enough, because technology touches every part of a company. Boards are adding cybersecurity and digital skills to their skills matrix, and she expects AI to be added too.
Karaboutis says adoption varies, with tech and software leading, followed by financial services and healthcare. Regulated sectors, governments and agriculture are slower. Inside companies she sees roughly 10% experts, 40% experimenters and 10% resisters.
The board sets and checks the strategy and makes sure risk stays within the company's risk appetite, meaning the amount of risk it accepts. Management runs the work. She sums it up as 'noses in, fingers out'.
Karaboutis says a ban pushes employees to use the tools outside the company, which can expose confidential data, as she says happened at Samsung. She prefers a secure in-house version and clear communication about what is allowed.
They use the same measures as any big investment, such as return on investment and key performance indicators, agreed at the start. Karaboutis adds that softer outcomes like employee satisfaction or building skills can count too.
Karaboutis advises dropping technical jargon and giving just enough detail on benefits, risks and dollars for the board to do its job. She suggests analogies, such as explaining layered security with locks, a safe and jewels, which she says is not dumbing it down.
Stated by the speakers. Each quote matches the transcript, which shows what was said, not that it is true.
Adriana Karaboutis says the technology and software industry leads enterprise AI adoption, using virtual assistants, chatbots and AI-driven cloud services.
“Tech and software clearly leading the path. It's part of the fabric of what they're doing. Virtual assistants, chatbots, ai driven cloud services, et cetera.”
Karaboutis says financial services is also a leading adopter of AI, using it for fraud detection, risk management and checking algorithms.
“Financial services, another industry way up there relative to their adoption around fraud, risk management, using it to check algorithms.”
Karaboutis describes a rough split inside companies of 10% AI experts, 40% experimenters or novices, and 10% who resist adoption.
“you still have about a 1040 4010 split, where it's 10 experts, 40 kind of people that are experimenting or in the novice phase, and then 10 that are actually”
In a survey by an organization Karaboutis calls North America Corporate Directors, tech, AI and cyber ranked in the top six trends directors worry about.
“Tech, AI and cyber were in the top six, along with geopolitical climate, et cetera.”
Karaboutis cites Samsung as an example where employees took company data into outside AI tools to create code or get insights, exposing potentially confidential information.
“We did see it some years ago and with Samsung, you know, hopefully done inadvertently, where people take data to the outside into these tools to either create code or get insights and then bring it back in.”
Karaboutis says that at a former company she imposed a lockdown of ten to fifteen days on AI tools while the company brought in its own instance, after telling the workforce.
“I said we're gonna lock down for a period of ten or fifteen days. But we let the workforce know this is coming.”
Michael Krigsman reports that Sam Altman said every successful startup founder can explain what they do in 25 words or less.
“he said that among startup founders, one consistent trait among every single successful founder is the ability to summarize and express what they're doing in 25 words or less in a really clear way.”
Karaboutis says boards use a skills matrix to assess which capabilities, such as financial expertise, they need among directors.
“Board of directors have what's called a skills matrix, and I would say all public company and I think private company as well, boards have these.”
Karaboutis says boards' skills matrices increasingly include cybersecurity and digital transformation, and she expects them to add AI.
“On those skills matrix, more and more we are saying cybersecurity, digital transformation, et cetera.”
Karaboutis says she joined the advisory board of an AI company, which she names as icreantree ai, that helps utilities, and that she holds a computer science degree.
“I myself have joined, you know, the advisory board, awesome company called icreantree ai. It's an artificial intelligence company that helps, even though I have a computer science degree and I'm tech savvy.”
Karaboutis says she has served on boards since 2015.
“And by the way, I've been doing boards for ten years, so it wasn't sort of I stopped full time and went into boards. I've been doing them since 2015.”
Karaboutis describes her move from operational roles to board work as 'rewiring' rather than retiring, a phrase she credits to a friend.
“When I stepped down from my last operational role, people said, Are you retiring? and I said, No, I'm rewiring”
Karaboutis says she believes Nestle has an AI ethics group, as an example of a company applying lines of defense to AI.
“Some big organizations, I believe a company, uh, Nestle, I believe, does this, where they have an F AI ethics group.”
Karaboutis says boards are trending toward applying the three-lines-of-defense model used for financial controls to AI oversight.
“We have the three stages, first, second, third line of threat vectors of, you know, financial controls. Well, it's this. It should be the same, and I would say is trending in that direction of three lines of defense around AI.”
Karaboutis says software and technology companies such as Microsoft and Google are rushing to use AI for competitive advantage.
“Software tech companies, I think, are rushing for advantage, no question. We're taking a look, whether it's the Microsofts, the the Googles, et cetera”
Karaboutis says boards do not ask ad hoc questions about headcount but instead check whether the company has enough specialists to achieve its strategy.
“Boards will not go in and say, um, you know, just, uh, uh, uh, sort of ad hoc. How many AI specialists do you have?”
Karaboutis says a board hires and fires only the chief executive, not the rest of the management team.
“We don't go, except for the chief executive, we don't hire and, and, you know, sort of fire the, you know, management team, et cetera.”
Karaboutis uses a locked front door, a safe and jewels as an analogy to explain defense in depth in cybersecurity to boards.
“I used to say, if you put a lock on the front door, but the jewels are sitting on the table, then not good, so maybe you have a safe”
Karaboutis says boards monitor the tension between AI's energy use and their ESG commitments but do not measure AI compute use against offsets one for one.
“As with any new technology, Lisa, there is what I call swings and roundabouts.”
Karaboutis notes her experience includes senior roles at Dell, Biogen and National Grid before becoming an independent board member.
“I've been doing boards for ten years”
Karaboutis surveys which industries lead and lag, and the split of experts, experimenters and resisters inside companies.
“Tech and software clearly leading the path. It's part of the fabric of what they're doing.”
Competitive pressure and efficiency push adoption, while skills gaps, ethics, privacy and compliance slow it.
“The drivers is around competitive pressures”
Karaboutis explains the board's governance role and why it must neither rush ahead nor lock everything down.
“Nobody ever cost, cut or risk avoided their way to prosperity.”
She argues companies should enable AI safely and communicate clearly, so staff do not leak data through outside tools.
“So organizations need to say yes and how versus no, because, right?”
Boards want education without jargon, and judge AI by standard return and performance measures.
“please start me with the forest before you take me to photosynthesis, right?”
She distinguishes oversight from execution and argues tech savvy should be spread across the whole board.
“it takes, you know, I'll use the phrase, it takes a village.”
Boards use audits and ethics groups for AI, and ask what it would take to reach outcomes, always within risk appetite.
“Ethics trumps profits.”
She advises tech leaders to learn the business, drop jargon, and help managers understand why boards ask questions.
“Recognize there's no recipe.”
How are boards of directors approaching enterprise AI and AI adoption in the enterprise, and what should boards do about members who don't understand technology? Let's get into it on CXOTalk number 866 with Adriana Karaboutis. She's an independent board member across multiple industries who has also held senior leadership roles at companies like Dell, Biogen and National Grid. Where are we in terms of adoption of AI in the enterprise? What are what are you seeing right now?
It is different by industry and different by company, but one thing that is absolutely consistent, it is permeating and growing by the day. Current state of adoption, I would say, is different by industry. Tech and software clearly leading the path. It's part of the fabric of what they're doing. Virtual assistants, chatbots, ai driven cloud services, et cetera. Financial services, another industry way up there relative to their adoption around fraud, risk management, using it to check algorithms.
Healthcare being pressured and using it quite a bit. And then you see some of the and you also have retail and manufacturing, and then you see some of the. I wouldn't call them laggards, michael, but I would call them a little bit slower to adopt your regulatory environments, your governments. Um, and you know, maybe some of the agricultural components, uh, or industries in agriculture, um, in terms of adoption inside the companies. In other words, as these companies and industries are trying to drive AI and really become consumers of it internally, you still have about a 1040 4010 split, where it's 10 experts, 40 kind of people that are experimenting or in the novice phase, and then 10 that are actually
still very reticent to adopt. They're the laggards. Oh, we shouldn't use this, shut it down, et cetera. Some of that coming from lack of education or or understanding of it, I would say, and some of it just coming from an abundance of being overly conservative and cautious. So there's quite a spread, but one thing's for sure, and that's what I said at the start. The the adoption is going faster and faster and faster. We're moving quickly, Michael.
When we talk about large organizations, what do you see as the drivers of AI adoption and also the obstacles or the roadblocks?
The drivers is around competitive pressures, companies, especially but public and private companies, I shouldn't say especially public. But companies are always looking to do more with less, to have higher quality, higher cost, customer satisfaction, employee satisfaction, returns for shareholders. So all the stakeholders are looking for better, more, um, improved, et cetera. So those pressures that cause you to look at tools and capabilities that will help you achieve that, and so that's some of the drivers. And as peers in industries adopt more of these tools and are able to actually show more capability and more of the outcomes that we're looking for, then that drives competitive pressures to all the companies in that industry to stay abreast and continue to improve.
Um, some of it is around newer products and offerings, some of it is around, you know, getting a better bottom line, which is cost efficiency, reduced s g and a costs, et cetera. Some of the roadblocks is around it is change, and there is just human nature and reticence to adopt, Michael. Um, lack of skills and understanding of the tools and technology, risk aversion, and boards in particular, have a big, big role to play around risk management and making sure that, um, the company is within what we call risk appetite, and that anything that is adopted or happens stays within that risk appetite.
Whether you're talking about cyber et cetera, and in particular with artificial intelligence and tools, there's ethical issues, there's privacy, there's security, um, there's ensuring there's no bias and fairness, et cetera. Regulatory compliance, legal compliance, all of those things are huge. And, you know, could be seen as roadblocks to overcome, but certainly slow, slow us down in our thinking and sort of our foray into artificial intelligence.
Andy, you talk a lot about risk. Of course, AI is all about innovation, and innovation implies something new that's changing and that's different, which, of course, is risk. And so, as a board member, how do you see this balance?
Boards are around governance and ensuring the company has the right strategy. Management teams do day-to-day operations and execute the strategy. So, as board members, we concern ourselves with do we have the right strategy in order to meet, um, our stakeholder commitments and the outcomes that we want, right? And so we we have a unique position. It's not overlapping with management, it augments management. There's a trust and a relationship and a and a and a and a collaboration that needs to happen with management.
Many have heard the expression noses in and fingers out. That sort of is the layman's expression for a board. In other words, we ensure we govern, we oversight, et cetera, and we ensure that risk management is, as I said before, within, um, an appetite level that we that that we set for the company. Um, so in that sense, boards increasingly recognize the impact of AI and everything that we do should be in line with our business strategy, our regulatory requirements, our legal requirements, et cetera.
And so we always make sure that how a management team is going to execute to deliver a strategy is within those parameters.
Can you dive into that tension between the desire to make use of that technology to achieve the organizational objectives and strategy versus the restraining factors due to the risks?
I'll use a comparison that should make it clear. Sports teams don't just play offensive plays, they have defensive plays. So in order to win the game, you go on offense and defense, right? Offenses we try things new. We look to technology to help us meet or exceed our business stated goals and objectives. And, interestingly enough, when board directors, uh, were asked by an organization called North America Corporate Directors, what are the trends they worry about most?
Tech, AI and cyber were in the top six, along with geopolitical climate, et cetera. So, given that, when we think about that, we we worry about it in a defensive play to ensure that, as I said before, we're staying in position in order to stay within risk appetite. We also want to leverage those tools, and we wanna see our companies and management leveraging the tools to accelerate andor meet, um, the business strategy. So there's the offensive and the defensive.
It's 360 degree wide angle lens, and it's critically important that boards don't fall into one, which could. If it's offensive, you could end up with a lot of privacy, regulatory, et cetera issues, or the other, where we're trying to preserve our way. We lock everything out, we don't let we say we don't wanna bring, for example, generative ai tools into the company, or machine learning, artificial intelligence, because we don't wanna run the risk.
Nobody ever cost, cut or risk avoided their way to prosperity. So there is that balance, and informed decision-making is critical, making sure that parameters are understood for. How is a company using AI? Do they have an, uh, a group that says we're looking at the use cases for what we wanna do? What are the risks of the use cases, and what are the benefits of the use cases? Those are very important things for management and in large-scale applications for boards to worry about as well.
So it's that balanced, informed decision-making, offensive and defensive plays in order to win the game.
Please subscribe to the CXOTalk newsletter, just go to CXOTalk dot com, subscribe to our YouTube channel. We really have extraordinary shows that are coming up, great discussions, and you should be a part of it, folks. You know, to some extent, it sounds like the, uh, the time of the beginning of social media. There needs to be a a kind of more thoughtful, in-depth approach that addresses the reality of how people live and work.
Technology is ubiquitous. We've heard that before. It is, um, we're led by what individuals do. I would say a great proportion of the population now, at least in the U.S., are using, you know, chat, GPT, DALL-E. We've seen students at school, you know, u- using this to write papers, et cetera. The exposure is there. Therefore, when people go to the office, we don't turn that off, right? And we look for tools enabled by the tech organization, et cetera, to be able to help deliver the outcomes.
I fundamentally believe people always wanna do the right thing, but they wanna be helpful and they wanna be productive, and they wanna deliver the outcomes. Or, I should say, and they want to be helpful, productive, and deliver outcomes. These tools enable that. So organizations need to say yes and how versus no, because, right? And so how do you? And again, it goes back to my offensive defensive play. But within these large organizations, how do we deliver the capabilities to use generative AI to create new content, right?
Text to photo to create, you know, illustrations that could help new products, new services. So unfortunately, regulatory lags behind sometimes the infrastructure of these large companies, which is important, by the way, to be able to renew and refresh. So that you've got good, clean, concise data- not concise, but good, clean, um, data accessible to be there. All of that has to be in place, and so we need to think about these tools are here, they are coming in.
If we don't enable them within the enterprise, you know, the worst of all worlds is where people take data. And I'm not saying anybody's done this, though. We did see it some years ago and with Samsung, you know, hopefully done inadvertently, where people take data to the outside into these tools to either create code or get insights and then bring it back in. You have now exposed potentially company secure, secure information and confidential information to the outside.
So it has to be. How do we bring in, not how do we create steel walls so people can't access these tools? You know, for a period of ten or twenty days, I mean, when I was at, uh, one of my former companies, I said we're gonna lock down for a period of ten or fifteen days. But we let the workforce know this is coming. We just need to bring in our own instance to ensure that we keep the security and the safety in.
And by the way, that is a really important point, Michael. Communication on how, when, how far, which data is available, et cetera, is super important, cause absent communication. Um, employees may not know what they can or can't do, or what they should or shouldn't do, and so all of that is critically important.
I really like the not defaulting to no for so many years, enterprise technologists, CIOs just simply the default answer. If somebody wants something new, is well, no, we can't do it. And then later, you know, we can take a look and see, well, maybe we can do it, but but the default is no. And so I really like your approach. Why don't we jump to some questions from LinkedIn and Twitter? There's a bunch that are coming in, and I encourage everybody who's watching, ask your questions, because literally, when else will you be able to ask somebody like Andy Karaboudis pretty much whatever you want?
So take advantage of it, okay? So the first question comes from Isaac Sokolik, who actually will be a guest coming up very soon on CXOTalk. And he says, Where are the board's mindsets today on AI? Is it a gold rush for competitive advantage, or are they looking for pragmatic ROI, or how are boards relating to AI?
Software tech companies, I think, are rushing for advantage, no question. We're taking a look, whether it's the Microsofts, the the Googles, et cetera, and and beyond, uh, they're really taking a look at how they can leverage AI for competitive advantage. And I would imagine that the board conversations are full of discussion around the opportunity, um, as well as the risk. But on that topic, um, as you move into what I would call the more traditional businesses, I think the discussions are still happening.
And some board members are quite bullish around. What are the advantages, what can we use, how can it help us? That's the big thing. How does it help us? Top line, bottom line, customer sat, employee satisfaction, you know, job availability, et cetera. But it's an intellectual curiosity that is actually something that I think is necessary of any board member, and so the conversations are there. Most boards are asking for more training in some of the more traditional businesses, training and understanding.
Now they don't need to understand. You know how you ingest data and all of the bits and bytes, but they do need to understand the fundamentals around. You know, what are the prerequisites? As I said before, your infrastructure needs to be simplified, your data needs to be available. And, by the way, this is a huge opportunity for the, you know, technical CIOs, CTOs in the room. Figure out and practice your game on how to communicate, because boards are asking for the education.
Please don't bury them. Stroke us in a bunch of tech speak around neural networks, et cetera. Practice your game to really be able to explain these. The the the technology, the capability, the risks, and how it could help your specific company and industry. That may be the difference between true success and what I'll call quicks faster success and PR versus a prolonged, um, sort of realization of value. So, um, Isaac, it varies, is what I would say.
I really want to emphasize a point you just made around the importance of simple, straightforward, clear communication and how important that is.
Absolutely, Michael, you know, my teams would would recognize and probably smile at the next phrase I'm gonna make. I always say, please start me with the forest before you take me to photosynthesis, right? So understand the big umbrella, even if when it's explaining artificial intelligence, it's the overarching umbrella. And then there's machine learning, and there's, you know, all of the categories there, generative AI, agentic AI, et cetera. And so when you take a board that is noses and fingers out, that has to really figure out and, um, synthesize a lot of information, comprehend it and come into insight.
They need to. You have to ask yourself, what do they need to know, and you know, what do they need? What do they need for informed decision making? And so that's a big capability that, you know, I'll put my CIOCTO hat on. We as technology people need to work on in order to help the boards be supportive and be able to balance that offensive, you know, with defensive play.
You know, I just this morning saw a segment from an interview with Sam Altman, the founder, CEO of OpenAI. And he said that among startup founders, one consistent trait among every single successful founder is the ability to summarize and express what they're doing in 25 words or less in a really clear way.
Absolutely right, you know, call it the elevator pitch, call it the you know, in 10 seconds, you've lost my attention, call, you know, whatever it is, but be very clear. And you know, there's the old adage of what do I wanna say versus what do others need to understand, and by the way, who are they? That's super important. So, um, I can't emphasize enough in E- And it's with any big breakthrough tech or any big concept, even if it isn't tech, you have to be able to really communicate.
What are we trying to do? You know, how are we gonna do it? And then I'll quote Simon Sinek and connect people to the why, right? And so that doesn't, that doesn't change for board members either.
Let's jump to the next question from LinkedIn and keep your questions coming on LinkedIn and on Twitter. And this is from Deepak Adina Rayana, and Deepak says, How do you measure the success of AI initiatives at the board level, and what metrics or indicators do you rely on to ensure these technologies are delivering both value and alignment to the company's long-term vision?
As with any initiative, major program or effort or transformation that we do at, um, you know, a major bigger, bigger, small companies, there is an expected outcome. And when you embark on something like this, which is a huge investment for the company from. And I don't just mean, you know, monetary, but resource, culture change, um, effort, discomfort, et cetera, there's an outcome that's expected. So measures are very simple, tried and true return on investment, return on equity, return for what we said we're gonna do.
How do we measure the successes? We're going key performance indicators. I know these are not, you know, big and sexy and different, but they're fundamentally. When we're doing something that's that big of an investment, what is the return we expect, and are we getting it in the timeframe in which we expected, right? So being able to measure your key performance indicators, your outcomes and output, the output steps to get to your outcome, et cetera. None of those things have changed.
I wish I could give a a fancier answer than that, but fundamentally, that's what we look for. But very important to be clear at the outset, what are we getting for this? And sometimes there are hard measures, most often, and sometimes there's also soft measures or soft outcomes. I would say, improved satisfaction from employees, those are real, they're important. So if we state them and do what we say and say what we do, that's where we m look at these outcomes.
And sometimes an outcome could be learning and saying only, as I said earlier, ten of the organization is expert. We wanna get to 20, that's an outcome. So it might not be monetary, but it is, you know, along the scale of what we said we wanna do in order to adopt something that could then provide stronger returns.
And this is from Cheryl Falk's Bendy, and she has a question about this relationship between management and the board, especially when it comes to AI. And she says this, following on Deepak's question, Are you even measuring AI initiatives discreetly, or leaving that to the management team and sticking to business outcomes that AI might be a component of?
When the initiatives are big and the investment is significant, or, once again, it's a significant change. Like we're trying to drive a different culture, organizational, a de, you know, uh, uh, adoption of AI. The board will get involved and we will take a look at it. And, you know, overall, we look at the maturity of artificial intelligence. What we don't do is go in and look at every artificial intelligence, um, initiative and say, how are each of these going percentage wise?
And things like that? That's for management to do again. I can't emphasize enough day-to-day operation and execution of initiatives to deliver on the strategy is the the job of management. Uh, the board assesses the risk, the board assesses the progress to commitments, um, and the outcomes, et cetera, as it relates to sort of, you know, the big overarching business strategy of the company.
It's a clear distinction, but it can also be murky at times. And and how do you decide when you need to kind of drill in more? Because at some point, you know, the board member may have to get involved. You may be able to offer advice or, uh, you know, there could be many, many different reasons. How do you balance that? how do you make that choice?
I get asked a lot about, you know, when do you know the difference and when to do it? It depends by company and by board, right? And it depends, it's steered by risk, right? So if it's not big enough to, um, uh, for a board to to worry about in comparison to other things on the agenda or happening, it's a very subjective. When is it time? You know, there a lot of people ask me, how often should the board, uh, get a review of, um, AI initiatives?
I don't know. It depends on. Are those AI initiatives integral and core to achieving the business strategy? Um, on the table, if that's the case, and there's a timeframe and a dollar amount and an importance level to it. Because it's it's the pivot point that would that needs to be done to get us there. Then I would say the board visits it quite a bit. But if it's not, and it's sort of, then we could get it in a pre-read, or we could get it just in an update, or it could be infrequent, so it depends.
Um, I wish I, you could say there was a recipe behind it, but there isn't.
Yeah, I can imagine there. These are complex issues that will vary according to the company culture and the situation.
That's right, in general, we're TR- cause we do track, are we staying abreast of technological advancements? So even if there isn't a specific initiative, we, we do. We need to ask leading questions like, how is the adoption of AI tools? what are we using it for? Um, what are some of the use cases that we're going after? Is it inventory management, for example, in, you know, in a retail space? Um, are we using it to monitor manufac- manufacturing quality and adaptive learning?
Certainly, if you're in the auto industry, the autonomous vehicle, I would say those boards are probably, and I'm not on a automotive board, so it's an opinion, um, but I would say those boards are probably talking quite a bit about the artificial intelligence. That is, you know, self-learning around autonomous driving if they have autonomous vehicles. So, again, it's it is a very, um, subjective to the industry and the company.
This is from Arslan Khan. He's a regular listener to CXOTalk, and he says this, he says, as tech becomes more and more a part of our lives, do boards need a tech advisor only, or a permanent tech board member? Who is fit to do this? The CIO, the chief digital officer, the CTO. So if you can address that, and I think it also gets to the larger issue of technology capability on the board.
Board of directors have what's called a skills matrix, and I would say all public company and I think private company as well, boards have these. And a skills matrix is around, um, a company assesses what do they need, what are the capabilities they need in their board. So, for example, um, financial literacy, financial expertise is one of the metrics, right? It may be sales and marketing, it may be product development, et cetera. On those skills matrix, more and more we are saying cybersecurity, digital transformation, et cetera.
I suspect we're going to start saying tech savvy, artificial intelligence and a a deepening of that. To have a token member on a board that says there is our digital tech person is a start, but it's not sufficient in my opinion, because technology impacts every part of companies. Whether it's from, you know, customer satisfaction measurement to, um, product development to, it is ubiquitous. So, in my opinion, we have to have, um, tech savvy board members, sort of permeating the board.
That doesn't mean that they need to know how to do microprocess, you know, programming, or that they need to, you know, do a Python program. But what it does mean is they need to be abreast, we need to be abreast of what are the technological advancements, how do they disrupt, how do they augment, what risks do they pose, what opportunities and benefits for our company and industry. And that broad-based technology, as I call it, I think, is imperative, if not for every member of the board, for majority members of the board.
And, you know, it's at one point, many years ago, on another board, someone said, Well, thank goodness we have Andy on this board, cause she's watching out for that. Well, I'm not sure that that is, you know, it, may maybe back then, but it's certainly not the right mindset now. So tech savvy is really important. And again, I'm gonna reiterate not everybody needs to go get a degree in computer science or computer engineering, but being abreast, staying in tune, intellectually curious, learning, et cetera.
I myself have joined, you know, the advisory board, awesome company called icreantree ai. It's an artificial intelligence company that helps, even though I have a computer science degree and I'm tech savvy. But, you know, it's out there helping augment human intelligence for utilities industries. That's how I stay abreast, so even people that have the degree can be not tech savvy. So it takes, you know, I'll use the phrase, it takes a village. That's how I feel about it.
What about the stereotype of board members who are old school, have very little comfort or understanding of technology, who are used to having secretaries, maybe in the past, who don't have the background? What do, what do you do if you're a board and you have esteemed members who are in that situation?
I'm an optimist and a glass half full. But I, I would say they're retiring because the pressures are such that you can't really. And I'm honored and pleased to be on the boards I'm on, cause I hand on heart don't see that. But I think they're retiring because they y- you just cannot, um, be an advocate for and do your duty of loyalty and duty of care that you need to do as a board member. If you refuse to sort of stay abreast of your inner industry and the technology that could impact it,
All right, so that problem is a problem that is taking care of itself over time, essentially. that's my view. Yes, This would be an excellent time to subscribe to the CXOTalk newsletter. Go to CXOTalk dot com and subscribe to the newsletter, and we'll notify you about upcoming events just like this, which we do all the time. So this question is from Chris Peterson, he's another regular listener, and he says, For those mature AI adopters, how do they measure positive and negative impacts of the tool and product choices that they make?
tools and product choices are not a decision for the board. So I will just say right off the bat, the board may ask questions of, you know, why did you bring an instance of chatgpt in or DALL-E or or? But it it is not a question for the board. As for management and what they choose to adopt, I, you know, I can assume, um, and believe that they make that determination based on the benefits, the availability, the costs, for sure.
And that is something, you know, boards will ask is how much we're allocating and and the cost of of AI, um, or tools, uh, the risks, and, uh, the openness. Is it open source or is it not? So for companies that are doing that, and I apologize, I'm, you know, sort of going between what the board would do and what management would do. Um, I assume they're doing, you know, balancing the equation as we did whenever we bought other kinds of tools around all of those parameters commercial relationship, viability, durability, um, you know, as I said before, price, performance, et cetera.
Here's a question from Lisa Pratico, and she says, AI used everywhere is showing a negative impact on sustainability with energy and water consumption. Are boards, or to what degree, are boards concerned about balancing the use of AI against, uh, negative ESG outcomes, or to ensure that that there are positive ESG outcomes?
As with any new technology, Lisa, there is what I call swings and roundabouts. And so boards recognize that while we have our ESG commitments and they remain strong, um, and we want to adopt the new technology, we have to monitor and understand both. And there may be a time where one is, you know, sort of, you know, usurping the other, but we monitor and we see growing pains in in this technology, um, is inevitable. And so we, I can't say that we sit and measure.
This is how much AI and compute power we're using, and this is the offset. But we are, in general terms, watching that f- as board directors, because we have ESG commitments and we have commitments on how to improve and do more with with less and provide better offerings. So it's not a one for one, but I think in the round, it will all come back to. We have a wide-angle lens and 360 degree view on these things.
Andy, as a board member, you don't directly have your hand on the rudder of a ship. You are giving guidance, and then there are then management is actually steering the ship in a in a more direct way. Is that frustrating for you at all? The fact that that you have this indirect role in terms of what the organization does as indirect role, as opposed to a very direct leadership management role?
When I stepped down from my last operational role, people said, Are you retiring? and I said, No, I'm rewiring, um, and good friend of mine actually coined that phrase, so I won't take credit for it. And when you rewire, you realize that you're going into a different set of roles. And by the way, I've been doing boards for ten years, so it wasn't sort of I stopped full time and went into boards. I've been doing them since 2015.
Um, but you, it's a different role, and with every new role, you ask yourself, what is my accountability, what is my responsibility, what is it that I'm bound to? Uh, you heard me say duty of loyalty, duty of care, et cetera. And how do I best do that to show up well to my boards? Now I will tell you, I went through growing pains. I would've been quoted as saying, if I could go back to my first board, I would give them all an apology.
Because it was that transition of management where I wanted to know, why are you doing this? et cetera, to a, you know, a board position. I'm gonna use your bo- your boat analogy. It was an excellent one. Management is driving the boat, but the board is overseeing, and the board doesn't come in, you know, way too late to find out. We thought management was in the Atlantic, and they're off in the Pacific Ocean, right? It could be more, it could be slight course corrections.
What the questions that we ask and the oversight that we provide are, are we on track? For, you know, have we delivered? Is the business strategy? Are the financials the way that we expected them coming in? And we ensure that we're delivering to the commitments made, and we help course correct. Whether it's succession planning, whether it's, um, you know, uh, uh, looking at the allocation of spend, whether it's, we have to look at competitive pressures and things that have changed, therefore, maybe we need to do a course correction.
So there is the pilot that is piloting or the captain that is driving, but then there's also the control tower that is monitoring, and so it the control tower plays its position, knows what it does, and has to synthesize a lot of information quickly to be able to be helpful and stay in the role that they're in.
So staying in the role is a key part of success for a for a board and individual board members. Mario Garcia asks, He says, AI is a game changer, but aligning strategy with responsibility, that's the real challenge. What do you think companies are missing?
first of all, understanding the true power of AI. So education, I don't know that they're missing, but certainly this is what companies need, um, to truly understand artificial intelligence capability and tools and offerings from software and tech companies and what they could provide. Clear tying of initiatives to the outcomes and the strategy of the company and what it's trying to achieve. Uber important measuring and ensuring that you don't have to course correct, cause course correction is good.
Staying with a strategy or with a tool or a capability that isn't working is just throwing good money after bad, and sunk cost is sunk cost. So there's a clarity of your strategy, the deliverables to get to your strategy, the outcomes that must happen, the tools and the and the programs that are going to help deliver that, and what I call stringing the thread through all of that. I think if there's anything, and I'll go back to your direct question of what companies might be missing.
It's really understanding the capabilities of AI, since, I mean, uh, given that topic, and what can or cannot provide for the achievement of the business strategy of of the organization.
Let's jump to another question this time, uh, again on LinkedIn from Demetrios Brempos, and he says, and this is a really good question, he says. Do board members face the AI change as an upskilling necessity, or as a more holistic culture change?
It is an upskilling, as I said before. I think as boards are evolving and sort of realizing the dramatic change that tech, cyber and AI are are are bringing, that board skills, board um, awareness and education need to, um, also, you know, sort of move commensurately. It is also a big culture change, it is a culture change within the company. Technology is now not just for programmers, right? I'm not gonna make a bold statement and say you never need programmers anymore, but you are moving further from that direction and away from the direction of programmers to.
Anybody can be a citizen developer, getting the output that they need, or or more people can. I don't want to offend anybody here on viewership, so I think it's both. It's the composition of the focus of the board, it's the composition and focus, education and management team, and it is a big culture change across the spectrum, from employees all the way to the board.
And this is a question from Lisbeth Shaw, who says, How are boards helping their companies address the ethical use of AI and trustworthy AI? Especially since doing so may cut into maximizing profit.
I'm very proud, and I'm not, you know, not just the boards that I'm on, but but most boards. Um, e- ethics trumps profits, right? I will just say that it's. It is a darn shame when we find the one or two unicorns that don't, you know, go operate that way. Ethics trumps profits. So how do boards ensure ethical usage through audits, through reviews, through stakeholder interviews, through third-party assessments? It's, you know, we do this, you know, on many fronts around strategy, et cetera, um, around financial controls, you know, we have our internal audits.
We have the three stages, first, second, third line of threat vectors of, you know, financial controls. Well, it's this. It should be the same, and I would say is trending in that direction of three lines of defense around AI. Some big organizations, I believe a company, uh, Nestle, I believe, does this, where they have an F AI ethics group. So the board asks these questions of management and make sure that these um controls are in place.
Do we have an ethics committee? What's the latest audit? Is our data secure? You know, how do we know we don't have biases in these programs? By asking these questions and forcing sort of that the responses, that's how we ensure and make ourselves comfortable. And, you know, we look at, we look at, um, data associated with that. That, in fact, we're staying ethical.
Okay, and on a related question, uh, Justin Menon on LinkedIn raises again the issue of the critical role of board involvement in navigating roadblocks to AI adoption while maintaining an appropriate risk profile. So can you just elaborate a little bit on this point, which really is, how can boards ensure innovation while at the same time not taking the ostrich approach and hiding from change?
The easiest way is we ask the question, what will it take? What will it take to be able to adopt artificial intelligence and AI tools in order to achieve X outcome or Y outcome? And um, if and then the board tries to dedicate resource or helps management supports management in dedicating resource or, um, funds, capital allocation to, uh, remove the roadblocks to achieve the outcomes. Not open-ended, though. How do we get AI tools in? is not something the board would say, right?
But if, for example, we're saying we want in, you know, uh, to improve inventory management and reduce shrinkage, for example, in a retail company, we would say, what would it take to bring tools in to do that? Management needs to be prepared, and again, very much a two-way street, Michael, but management needs to be prepared with our infrastructure is legacy. Here's what we need. This is what it would take, um, or these are the tools that we need to invest in, or these are the resources that we need to go hire or get help with, right?
But it's how do we achieve, what are the roadblocks we can help remove, and then setting the support structure up for management to do that. But again, not just to bring in AI tools, but to achieve the outcome of what these tools could promise for the stakeholders.
Correct me if I'm wrong. It seems, in a sense, what you're saying is the board's responsibility is to drive the organizational agenda at the same time, balancing competing forces. Ultimately,
It's to agree to the organizational agenda and the strategy, um, of the company, and then to see how the board. What is needed from the board to support removing those roadblocks while constantly monitoring that we stay within risk appetite.
the way that you're describing it, if I can be totally blunt, it sounds like kind of corporate speak. That's pretty much what I said, or am I missing? what am I missing here?
You missed the risk appetite. Okay, all right, So you said the board will remove things, but the board will. There's a check and ba- look, I'll take the corporate speak out of it and just I love that you called that out, uh, Michael, but it, it, there's checks and balances, so it's not, yeah, go, it's again. I'll go back to what I said 40 minutes ago. There's an offensive play, but we have to make sure that we stay within our regulatory, legal, ethical, um, boundaries.
And I'll always repeat that. Sorry if that, yeah,
very helpful. So checks and balances is a very essential element of this.
absolutely, absolutely. So if I, if I may ask, as a board member, Andy, would you give us your view and opinion on, you know, chatGpt? And it's great, it's this, it'll do this, and what are we training? And we used, you know, 6 billion um, data sets and all the rest of it, that's fantastic. But in the same breath, I'm gonna say, how, you know, what are we doing to ensure that corporate data does not get out?
As we're, as we're delivering on this, can't emphasize it enough. I know I've said it about twelve times, so forgive me.
Well, it's clearly an essential part of the the board role, and so it's great that you're emphasizing it. Uh, we have a number of questions left. I'm gonna ask you to answer these really quickly because I want to have time to share advice on how, uh, technology leaders can become board members. So, very, very quickly from Twitter, from Arsalon Khan again, strategy without execution is hallucination, sometimes execution is a fraction of what strategy wanted. What do boards do?
We make sure that the plans for execution are sufficient and enough for to achieve the strategy. And by the way, some of those are not just tech plans or AI plans. So, y- this goes to threading the needle strategy, strategic, you know, uh, um, uh, initiatives to achieve that strategy, that outcome, execution being in line with those initiatives to achieve the strategy. I know I'm very sounding making it sound very simplistic. Quite often, companies have strategic goals, but don't have the right strategy to achieve them, which means they don't have the right execution, um, focus a and programs, but it's it's a thread the needle.
Let's go back to LinkedIn very quickly from again from Isaac Sokolik. Are boards diving into how effective management is in investing in leaders and employees for training around AI and effectiveness on change management, or is this left to be a management issue?
If we have set up the outcomes and the commitments and the strategy again that we're trying to achieve, there is a set of capabilities, programs, initiatives, efforts, culture changes, et cetera that are de- that are have been set forth to achieve that. You measure if those things are being achieved. Boards will not go in and say, um, you know, just, uh, uh, uh, sort of ad hoc. How many AI specialists do you have? We will say, do we have enough specialists to achieve this?
Do you have the workforce of the future? Are they located in the right places? Do they have the right reward mechanism? Are their objectives aligned? We will ask those questions. We don't go, except for the chief executive, we don't hire and, and, you know, sort of fire the, you know, management team, et cetera. We have a lot of input, but we make sure that companies are resourced just the same way. We make sure that capital allocation on a financial perspective is done, you know, sufficiently to achieve the outcomes that we're trying to achieve.
Again, you've got the the checks and balances to make sure that resources and people are in place who can carry through the the vision and the strategy of the organization.
It's exactly right again. Noses in, fingers out, sniffing it, ensuring that sort of thing. We don't execute.
We have a question from Cheryl Falk's Bendy again, who says, What do you look for in a company when you're considering joining a board?
First of all, is it an industry that I am interested in and can contribute to? Um, is it a company that I believe, um, I could am interested in and can contribute to and can benefit from my expertise? Is it a board that I can work with? I look at by name and background each board member to see is that a board that I can learn from and contribute to. Um, is it an ethical company? I look at the financials, you know, sometimes they're great, sometimes, you know, you, you can find opportunity to grow that, but that's not typically how I the, you know, a a key criteria for s for assessing it.
And so through that, you start realizing, is it a place where if I go play on that field, I can actually contribute? And, um, you know, give good guidance, good insight, and where my capabilities are needed. I consider it from my personal experience, I consider it a duty and so, um, and some, you know, an honor, if you will, to be on a board. And those are the things that I think about.
What advice do you have to business leaders inside an organization on being effective in working with the board?
So I have seen managements teams that try to manage the board, um, and to some degree, there is a level in that collaboration where you need to manage each other's expectations, et cetera. But for me, it's around collaboration. What does the board need to do to do their job? What does the board need to do to be effective at, you know, their oversight and governance, and to help give us guidance? That's what I would look at as in in a management team.
Quite often, we look at board members when they ask questions, Oh my gosh, the board asked for this, relax, right? Why is the board asking, what is the board looking for? And so, just like this is gonna sound very trite, Michael, but just like in a good relationship, you want, you know, someone asks you something. Don't go on the offensive or defensive, but what is it they're looking for? And how do I help achieve that so that they can help me through oversight to deliver that?
That's how I would view it.
So really understanding, trying to understand the board member's perspective, what do they need, and how can I, as a member of the leadership or the management of the company, help that board member achieve whatever it is that they're looking for?
That's right, and it's quite often it's fair to say, why do you ask? Because it could be something completely different than they what they might interpret. It could be just a curiosity, it could be some other other reason. Or the board member may say, you know, I, I can't really divulge, or whatnot. I just, you know, I, I, I need to understand this, this bit of information, but typically we go through CEO for that.
What advice do you have for technology leaders, CIOs, CTOs and so forth, who want to become a board member?
Recognize there's no recipe. So if someone says to you, go do this, go do this, go do this, right? I've seen that quite often. It is not a recipe. There is chemistry to bring people on boards. There is, um, experiences that the, you know, certain companies are looking for. I mentioned capability matrix, et cetera. But here is what I would recommend to every tech person figure out the business, understand. If you're going for a particular company or in an industry, start understanding the industry.
Your trade may be technology, and that's what you bring, but you're going on a board for governance, oversight, fiduciary responsibility, your, you know, financial literacy, et cetera. So hone those skills and b- bring yourself, you know, up sort of into that broader umbrella and lose the tech speak. Quite often, we are so impressed with ourselves on all of the deep technology that we know, and we should be very proud of it. But that is not the level that boards operate at, you know, occasionally your tech boards may want to go deep into things, but you really need to raise your game up a level
You mentioned this earlier. Can very quickly. Can you give us an example to make this point concrete? If I'm a CIA CIO and I'm talking with the board, how much technical detail should I provide?
enough. So that they get the picture of benefit, risk required, you know, dollar allocation, if that's what's being asked for, and so that they can do their job. So recognize the board's job, right, recognize their oversight and governance. What do they need to know in order to do their their job well, right? And so I use a lot, and I'll say something else. I use a lot of analogies and comparisons, right? You know, we were trying to describe threat vectors and you know how people can infiltrate, what the bad actors do, for example, on cyber.
And so, you know, draw some analogies. I used to say, if you put a lock on the front door, but the jewels are sitting on the table, then not good, so maybe you have a safe and then you have, you know, it under a, you know, the boards in the kitchen. That's defense in depth. So I would explain it in terms that they could. It's not dumbing it down. It's just relating it to something they know.
That's an interesting point. It's not dumbing it down, but it's presenting it within the context of what's important to them and where their focus lies. OK, well with that we are out of time, so I want to say an enormous thank you to Adriana Karaboutis, who is a multi industry board member. She's held senior leadership roles at large organizations like Dell, Biogen, and National Grid. Andy, thank you so much for being with us and sharing your expertise with us today.
Thank you, Michael. It's a pleasure being here. I appreciate you having me back.
And a huge thank you to everybody who asked such amazing questions and who were who's watching today. Before you go, please subscribe to the CXOTalk newsletter. Just go to CXOTalk dot com, subscribe to our YouTube channel. We really have extraordinary shows that are coming up, great discussions, and you should be a part of it, folks. Thank you so much, everybody. I hope you have a great day and we will see you again next time.