IBM Institute for

Business Value

CEO decision-making
in the age of AI
Act with intention

Global C-suite Series

28th Edition
The CEO Study

i
 CEO decision-making
in the age of AI
Table of contents

Key takeaways 3

Introduction 5
About the study

The IBM Institute for Business Value, in cooperation Part One: How CEOs make decisions 11
with Oxford Economics, interviewed 3,000 CEOs from
30+ countries and 24 industries as part of the 28th Part Two: Applying decisions across the organization 23
edition of the IBM C-suite Study series. These
conversations, completed from February through
Part Three: What decisions matter most? 31
April 2023, focused on executives’ perspectives on
leadership and business; their changing roles and
responsibilities; and CEO decision-making today, Part Four: Better decisions for a better world 39
including key challenges and opportunities, their use of
technology, data, and metrics, and their visions for the Action guide 42
future. Insights were also drawn from numerous client
conversations including more than a dozen deep-dive
Research and analysis methodology 44
interviews with CEOs conducted explicitly for this study
from March through May 2023.
Notes and sources 46

1
 Key takeaways

30+
CEOs are weighing unprecedented urgency and risk.
With a host of new decision-making inputs—from a growing amount of
data to a widening pool of increasingly involved stakeholders—it’s no
Countries wonder 44% of CEOs regret a public stand they’ve taken over the last
three years.

3,000+
Generative AI changes—well, everything.
Three out of four (75%) CEOs believe the organization with the most
advanced generative AI wins. Already, 43% of CEOs are using generative
AI to inform strategic decisions.
CEOs
CEOs feel their organizations are ready for generative

24
AI. Other executives are not as confident.
69% of CEOs see broad benefits of generative AI across the organization
but only 29% of their executive teams feel they have the in-house
expertise to adopt generative AI.

Industries
2 out of 3 CEOs are acting without a clear view
of how to help their workforce with the disruption
and inevitable transitions AI will bring.
Fewer than one in three CEOs say they have conducted an assessment
on the potential impact of generative AI on their workforce, yet are using
it already.

A lack of clarity is impeding decision-making and

investments.
More than half (56%) of CEOs say they are delaying at least one major
investment due to a lack of consistent standards. CEOs face a lack of
consistent standards, particularly in emerging areas such as
sustainability and data and privacy.

2 3
 Introduction

“Generative AI Any decision that makes its way to the CEO is one
that involves high degrees of uncertainty, nuance,

models surprise,
or outsize impact. If it was simple, someone else—
or something else—would do it. As the world grows

impress, and
more complex, so does the nature of the decisions
landing on a CEO’s desk.

scare us, all at It used to be easier: CEOs relied on primarily financial results and
operational reports, combined with intuition and experience, to

the same time.”

make decisions about the next best move for their organization.
That’s not enough anymore.

Ask CEOs about their most challenging issues now and top-of-list
items move far beyond financials into complex issues that require a
host of inputs—AI; sustainability; cybersecurity; diversity, equity, and
Gonzalo Gortázar inclusion; stakeholder management. CEOs tell us it’s almost crushing.
CEO, CaixaBank

It is an
epic challenge

4 5
 Enter generative AI Figure 1

Decision-making has become an epic challenge in And now, generative AI has burst into the scene— Pressure to accelerate
an environment where nuance and compromise a technological tool that promises, potentially,
AI adoption runs high
are seen as weaknesses. Today’s stakeholders to deliver answers to even the most vexing Pressure to accelerate
show no tolerance for missteps. Seemingly Pressure to slow
questions, instantaneously. With advanced
everyone from shareholders, employees, and the analytics and deep data, CEOs can now turn to
media hurl “Why that?” assaults no matter what artificial intelligence to gather insights and 66%
decision a leader makes. Yet, 77% of CEOs say direction. It may be tempting for enterprise Board members
they feel compelled to take a stand on potentially leaders to see generative AI as their magic 10%
controversial issues; 61% say the same even wand—enter a few prompts and seconds later
64%
when they don’t see clear financial gain for their have a compelling rationale that supports
Investors/creditors/lenders
organization. And it’s no wonder 44% of CEOs any decision. Except of course it doesn’t work
say they regret a public stand they’ve taken in 1%
that way. But that’s not stopping stakeholders
the last three years.1 Leading can feel like a from pressuring executives to use it. Investors are 55%
no-win situation. still the most heavily biased toward acceleration, Employees
but a slightly higher percentage of board
Intentionality matters more than ever. In this 28%
members are pressuring CEOs to accelerate
environment, advancing a strategic roadmap is also
adoption (see Figure 1). 55%
more essential than ever. It’s not action that matters Media or press
most, but output. And the right outputs depend on As CEOs respond to the pressure to accelerate
16%
the right decisions. AI adoption, their teams are more hesitant. CEOs
firmly believe in the benefits of generative AI 53%
across their organizations but other executives cite Government/regulators

As CEOs respond
a lack of in-house skills. CEOs (74%) agree or 15%
strongly agree that their team has the knowledge
49%
to the pressure and skills to incorporate new technologies such as
generative AI. Just 29% of other executives,
NGOs/advocacy groups

to accelerate AI though, agree that their organization already has 18%

the expertise in-house to adopt generative AI and
adoption, their teams only 30% agree that their organization is ready to
Business partners
49%

are more hesitant. adopt generative AI responsibly.

16%

48%
Customers
26%

Q: To what extent are the following stakeholder groups applying pressure to slow or accelerate adoption of generative AI?
Note: Respondents were not asked to indicate the direction of pressure if they identified a stakeholder group as applying
little to no pressure around generative AI.

6 7
 3,000 CEOs from around the globe share
their thoughts and experience
“The biggest feature of
To address this pivotal inflection point, the IBM
Institute for Business Value (IBM IBV) has
undertaken its most complex and far-reaching
Our analysis explores a few key themes:

How CEOs make

turbulent times is the
CEO study. We have surveyed more than 3,000 CEOs
and public sector leaders globally about their key
decisions increase in uncertainty.
A large part of growth
decisions and processes, what matters most now,
and what they see ahead in the next three years. We
determined what separates CEOs who are making
Applying decisions
across the organization
comes from dealing with
decisions most effectively from those who are
becoming increasingly bogged down by uncertainty.

that uncertainty.”
We also explored, with a select group of CEOs, how What decisions
generative AI is impacting decision-making not only matter most?
in the C-suite but throughout their organizations. We
talked in depth with CEOs across multiple industries
about what it’s really like to lead in this environment, Better decisions
as the situation unfolds.
for a better world Company Chairman, Asia Pacific

What follows is a roadmap for leaders, based on

conditions right now, on how their decision-making
must evolve and adapt. Across this analysis we delve into executive-level
strategies, as well as organization-wide adjustments
impacted by AI and other data-fueled tools.

None of us has a crystal ball, but it’s not simply

knowing the future that makes you successful. It’s
what you do with what you know.2 High-performing
CEOs are navigating not only what is possible to
know, but how to laser-focus on separating the signal
from the noise.

8 9
 The decision-making burden on
CEOs has always been heavy. Broader
digital accessibility and new tech tools
have unleashed a wave of innovation,
forcing leaders to make decisions and
act more quickly than ever to stay
ahead of the tide.

An explosion of real-time data supplements

traditional financial and operational metrics and
provides an ever-broadening array of information.
While often useful and occasionally instructive, the
swell of continuous inputs from so many sources
can become overwhelming. Yet, top CEOs
consistently use a range of decision-making inputs
and are more than 13x as likely as their peers to
rank their enterprise dashboard as excellent at
providing insight.

Part One

How
CEOs make
decisions
10 11
 External conditions have made CEO challenges
The majority of CEOs
61%
even more complex. From the COVID-19 pandemic
to climate change, increasing cyberattacks to believe the most advanced
shifting workforce expectations, CEOs face
of CEOs lack consistent
standards in one or more
decision-making that goes well beyond traditional generative AI wins
shareholder value models and their own personal
areas of strategic focus
experience. Unpredictable, asymmetrical situations
In many ways, one of the most critical decisions Yet CEOs must also weigh this “act-now” pressure
are arising more frequently, involving more factors
CEOs face right now is whether and how to use AI against potential dangers such as bias, ethics,
and more stakeholders. In Greek mythology,
—in their own decision-making and throughout their and safety. More than half (57%) of CEOs are
Hercules slayed Hydra, the many-headed serpent,
enterprises. With the public explosion of generative concerned about the security of data and 48%
but each head he removed was replaced by two
AI tools, the phrase “AI” has become dominant in worry about bias or data accuracy. CEOs wrestle
others.3 In similar fashion, CEO decisions center on
press accounts as well as quarterly earnings calls, with charging boldly into a still emerging space or
increasingly complex and fast-changing topics that
in a way no CEO can afford to ignore. Some acting with greater caution and risk being left
frequently involve an ethical or moral component—
companies have seen billions in valuations behind.
and deciding on one topic can cause other issues to
evaporate after mentioning the threat of AI; others
rise to the fore. There is nothing simple about this conundrum:
have pointed to AI as the holy grail for their future.4
how to address AI in all its manifestations—from
Adding to these challenges, CEOs face a lack of
The rush to action is understandable. Three- generative AI to intelligent automation—even as
while standards around decision-making, particularly in

56%
quarters of CEOs (75%) believe that the enterprise the tradeoff of benefits versus drawbacks is still
emerging areas such as sustainability and data and
with the most advanced generative AI will win and unfolding. As President and CEO of American
privacy. CEOs tell us common standards could help
they say competitive advantage depends upon it. As Honda Motor Co., Inc., Noriya Kaihara, puts
are currently delaying a smooth the way, speeding decision-making
they race toward AI superiority, 43% of CEOs say it, “So, the question we ask is: can we use the
major investment pending throughout the organization.
their enterprises are already using generative AI to technology to help people get things done?
greater clarity around
It’s no surprise that CEOs are drawn to inputs that inform strategic decisions, 36% for operational Can we use the technology to help people achieve
standards and regulations
promise to simplify or accelerate decision-making, decisions, and 50% are integrating it into their their potential? That’s what we are trying to
such as data-driven, AI-backed tools—and yet here, products and services. do. Technology should help make our business
too, the complications intensify. Generative AI, in simpler and easier. But, we are always
particular, when asked a question, often expresses thinking about the consequences of using
itself with certainty, but sometimes “hallucinates” each technology.”
its answer. AI trained on generic datasets—AI that
lacks transparency in its logic—could function as an
opaque engine of mistakes and misinformation,
hampering rather than helping CEO decision-
making. Because generative AI amplifies both the
good and the less-than-ideal that exists in any
organization, it needs to be applied in line with a
cohesive strategy rather than as a fix for a multitude
of ad hoc situations.
12
12 13
 Figure 2

What inputs do CEOs use

75% How CEOs decide to make decisions?
of CEOs believe that what inputs matter
competitive advantage will
depend on who has the most
advanced generative AI
CEOs draw from a variety of inputs when making
strategic decisions. An increase in the amount of 76%
data they must consider from newer areas like ESG, Operational data
added to a growing number of external inputs,
means there is just more to consider than ever
before. In our global survey, most of the CEOs report
that they still rely prominently on operational data 75%
(76%) and financial data (75%). Yet more than three Financial data
out of four CEOs stress that the most important

63%
decisions cannot be made on data alone. In fact,
63% of the CEOs turn to input from their people, and
over half (54%) include personal experience and
Internal input
intuition in the mix (see Figure 2).
Already,

50% 54%
Personal experience
of CEOs are
integrating generative
“So, the question is:
AI into products 50%
and services
can we use the Thought leadership

And 43% technology to help 42%

people get things done?”
are using generative AI to Other enterprise data

38%
inform strategy decisions

External input
Noriya Kaihara
President & CEO, American Honda Motor Co., Inc.
34%
ESG data

Q: How often do you rely on each of the following sources of

information when you make strategic decisions? (% reflects
responses for “Most of the time” or “Always”).

14 15
 Through our research and interviews, we sought
to understand what sets apart the CEOs who are most
First, these CEOs have high confidence in their
digital infrastructure and data, and they believe “Sometimes you must realize that
a situation cannot wait for more
successfully navigating today’s complex decisions. those capabilities enable better investments and
higher efficiency in delivering value. But they
When we looked at our high-performing group
also express a strong conviction that the most
through a financial lens, they deliver far better
results than their peers, with 2023-2025 projected
important decisions a CEO makes can’t be made
on data alone.
relevant data and take action
with what you have at hand. With
annual revenue growth 21% higher than the
average respondent and annual operating margins This group of decision-makers is more than twice as
24% higher. likely as their peers (65% versus 28%) to strongly

But two other key elements also rose to the top.

agree that metrics are driving organizational
behaviors, and they say metrics give them a full
the data that exists, you have to
understanding of organizational performance and
health. They have effective lines of communication
decide at that moment because
you can’t wait six months to have
with key stakeholders and are confident they are

“Decision-making based well-positioned going forward.

on intuition, common the right data set, or the right

sense, and knowledge tools, or the right process.”
is very good and should
never be lost. But the Fabián Hernández
CEO, Movistar Colombia

more analytic support

we have, the better.”
Gonzalo Gortázar
CEO, CaixaBank

16 17
 Case study

What’s perhaps most telling is how CEOs make CEO Fernando González of Mexico-based building
Pushing smart
strategic decisions amidst uncertainty. materials giant Cemex explains how he approaches
different types of decision-making: “The orthodox decision-making into
It’s not that data doesn’t matter. These CEOs
emphasize that clear metrics drive outcomes from
response is that decisions should be data-driven. the organization
In many situations, this works. You adjust plant
investor confidence to regulatory compliance to
operations based on data—for example, an oven
employee recruitment. But they are not looking
temperature should be based on data, not opinion.
solely at the numbers. In fact, they are far more
But in other situations, it is not as clear-cut. For
likely than CEOs overall to use a broad range of
example, deciding on an investment requires data
planning approaches, including forecasting and United Arab Emirates-based Majid Al Futtaim Retail is a franchisee
but also other variables. How much do I trust the
modeling, scenario-based planning, benchmarking, for French grocery retailer Carrefour, operating some 450
source of the information? Is the criteria behind the
and data mining. Lawrence Lam, CEO of Prudential Carrefour locations in 16 countries across the Middle East, Africa,
data correct? There are other reasons besides data
Hong Kong, describes his approach: “We make data and Asia, serving 750,000 customers daily and employing 37,000
behind some of my decisions—and I need to know
work for us, not vice versa. Our corporate focus is team members.
when enough data is enough.”
customer centricity so the right mindset about data
Challenge: Shopping patterns and purchasing behaviors are very different
is built on customer centricity—knowing what CEO Dirk Adelmann of German-based mobility firm
across Majid Al Futtaim’s store footprint. With a mix of consumers
customer pain points we want to solve, what data is smart Europe GmbH says he uses data but balances
representing many nationalities, the company struggled to match
relevant to drive that transformation, and where that approach by tapping into human experience
demand and store-item assortments. For analytics, Majid Al
that data is available or not.” and wisdom frequently: “If you start asking the right
Futtaim relied on an on-premises data warehouse solution that
questions, you will see that the ideas are coming
required significant manual intervention. SQL script coding was too
quite naturally from your teams. When I ask one
slow to stay ahead of growing data demands and market
person a question, two others are listening in and
“The orthodox response
complexity.
also have an answer. That is much more effective
than me always giving choices for the direction we Solution: Majid Al Futtaim’s leadership committed to more sophisticated
is that decisions should should go.” data-driven decision-making to enable faster real-time
adjustments. They implemented a hybrid cloud data and analytics
be data-driven. In many platform architecture, a unified data hub for advanced analytics,
and data-model development with built-in governance
situations, this works . . . capabilities.

in other situations, it is Results: Majid Al Futtaim is now able to make in-store changes faster and
more precisely. With more advanced analytics and less time spent

not as clear-cut.” managing data, the company can test scenarios with a fail-fast
mechanism, use data science tools to evaluate models, and
ultimately get to market more quickly. The turnaround time for
responding to business requests has improved by 100%.
Fernando González
CEO, Cemex

18 19
 Perspective Perspective

What is the top The external forces

barrier to AI adoption? impacting CEOs over time

When CEOs struggle over adopting generative AI, their first concern is The most important external forces impacting CEOs and their organizations have
data. Without trusted, reliable data, even the best AI will deliver faulty, varied since 2004, when the IBM IBV conducted its first CEO study. Technology
Biggest barriers to
biased, or dangerous results. Yet getting your data house in order steadily rose, peaking in 2012, and has remained fairly steady ever since as a top
generative AI adoption
is no small task, and for many enterprises one that is far from complete. factor. Workforce and skills have been more variable, but appear to be rebounding

61%
The top barriers: data lineage and provenance, a lack of proprietary data as an issue of growing importance. And regulatory factors have been on a fairly
to customize, and security concerns. steady climb since the study’s inception, clocking into second place this year.

Concerns around data Data concerns, however, reach beyond generative AI. Organizations have
lineage or provenance faced data challenges for years. In our research, top-performing Technology and regulation continue to significantly impact
companies in revenue, growth, and tech maturity focus on data standards organizations, while workforce and skills are rebounding

57%
and quality in ways that lagging peers fail to. The best CEOs know that,
while not glamorous, fixing data shortcomings is an essential priority for 48% | Technology factors
competitive advantage.
Concerns about 44% | Regulatory factors
security of data
Overall data challenges for the organization 43% | Market factors

53%
Constrained by
Unclear data calculation and
reporting across suppliers/partners
41% 36% | Workforce and skills

31% | Environmental issues

Difficulty identifying
regulation/compliance
40%
meaningful insights
30% | Socio-economic factors
Unclear data calculation and
reporting in your organization
40%
29% | Globalization
Inconsistent standards/formats/
frequencies across datasets
37% 22% | Geopolitical factors

Poor visibility into data sources 20% | Macro-economic factors

and quality
36%

20 20 20 20 20 20 20 20 20 20 20 20
04 05 08 10 12 13 15 17 19 21 22 23

Q: To what extent do you face the following data challenges today? (% reflects Q: Which of the following external forces will most impact your
responses of a “great” or “very great” extent). enterprise over the next 3 years?

20 21
 CEOs cannot just focus on the way
they make their own decisions. Good
Deciders-in-Chief also guide decision-
making across the organization.
They set a framework for how decisions
are made up and down the line, and
how information about those
decisions flows.

“If there are too many levels from the CEO to

the grassroots employees,” observes a chemicals
company chairman, “all the information that should
be transmitted to the upper level will be lost.” The
same intentionality that CEOs apply to their own
decision-making must be baked into the decision-
making systems across an enterprise.

“It is ultimately up to me to make sure that the

organization adopts analytics in the way that is
appropriate,” says CaixaBank CEO Gortázar.

Part Two

Applying
decisions across
the organization
22 23
 Managing AI from the bottom up 84%
Our research shows that employees are already using generative AI platforms in To get ahead of this wave, top CEOs are initiating of CEOs have generative
their work—whether that’s approved by executive leadership or not.5 Done within and deepening conversations with their teams AI use cases completed
a well-designed system, this can dramatically accelerate speed to insight and how about the use of AI—to both remove roadblocks to or under development in
quickly organizations can act on those insights. But it requires AI that “shows its progress and to ensure safety measures are in critical functions
work” to avoid misinformation—meaning AI that allows transparency into how place and promote responsible AI. They need

70%
it generated its insights—as well as a platform that enables secure use of guardrails that align with the organization’s values
proprietary information. Koichi Kameda, CEO of retail business Trial Holdings, and standards. They also need team members who
Inc., explains: “The company’s basic premise is to solve customers’ problems, have AI skills. “Acquiring digital experts is one of
and it intends to explore what technologies can be used for what purposes in of CEOs have
our biggest challenges, and I’m sure this is a
order to achieve this goal.” formal business plans
challenge for all industries,” says Tomoyuki
for generative AI
Takaya, CEO, Cardif Assurances Risques Divers
In all functional areas, generative AI use cases are ahead of formal business plans completed or under
Japan / Chief Marketing and Transformation
for generative AI. This trend indicates many companies are still in exploration development in critical
Officer, BNP Paribas Cardif Japan. “We need to be
mode. We asked CEOs about more than a dozen corporate functions and in every functions
more flexible to secure digital talents, for example,
category, at least three-quarters of CEOs expect to have at least one use case
through collaboration with our ecosystem partners
complete and ready to deploy within the next 12 months.
and other industry players.”
Creating rules around generative AI can optimize the benefits while protecting
CaixaBank’s Gortázar was frank about the mixed
ethics, avoiding bias, protecting intellectual property, and so on. Yet just one
emotions AI exploration is causing: “Generative AI
in four CEOs in our survey say they have issued any guidance on the use of
models surprise, impress, and scare us, all at the
generative AI within their organization. Some CEOs are now issuing moratoriums
same time.”
or bans on its internal use, particularly in industries such as financial services,
technology, and communications.6 What remains to be seen is where they go
from here—whether or not this is just a temporary pause while they determine
appropriate security, guardrails for use, and foundational models.

Meanwhile, teams are developing specific use cases–without an overarching

organizational plan. In manufacturing, for instance, 34% of organizations have
generative AI use cases ready, versus just 19% with formal plans.

24 25
“Effective decision-making is a Workforce planning Figure 3

in the hotseat CEOs make workforce

combination of data, human judgment, decisions due to generative AI
As AI invades the workplace, both consciously
and people’s opinion. The best decisions and surreptitiously, it is also fueling workforce
disruption. Our research shows 43% of CEOs say 46%

are those where collaboration informs

43%
they have reduced/redeployed their workforce due
to generative AI, with an additional 28% saying they

the process. 36%

plan to do so in the next 12 months. Yet a similar
proportion of CEOs report having hired additional
people due to AI, with plans for more hiring ahead.
The picture is muddled. Answering the question 28% 28%
So, we take input from diverse groups— “what kind of workforce will I need” seems to be a 26%
decision that is largely unresolved.

subject matter experts, data analysts, Yet fewer than one in three CEOs have assessed the

business leaders, and frontline

potential impact of generative AI on their workforce.
This is among the most disquieting findings from
our analysis. It means two out of three CEOs are
employees who interact with customers. acting without a clear view of how to help their
workforce with the disruption and inevitable

We take input from all these categories transitions AI will bring. Whether this is an oversight
or a process that is lagging, it is something to

before making a decision.”

watch—particularly as there appear to be significant
Hire additional Reduce/ Assess potential
gaps between CEOs and other senior executives
workforce redeploy impact on
when it comes to their sense of the organization’s workforce workforce
skills and readiness.

Baby George
Actions already taken
CEO, Joyalukkas Actions planned in the next 12 months

Q: Which of the following actions have you already taken/will you

take in next 12 months with respect to generative AI?

26 27
 Figure 4

CEOs are eager to act on AI, but

“Try to engage your Big questions also linger about the organizational
readiness to embrace and implement AI in decision-
their teams are less confident
team. Before making making. Seven in 10 CEOs see broad benefits from
AI across the organization but their executive teams

an important decision, I are far more wary about AI and workforce readiness
(see Figure 4). Only 29% of teams feel their 7 in 10 CEOs say generative AI . . . while executives cite a lack

try to confirm if my view organization already has the in-house expertise

to adopt AI.
brings broad benefits . . . of organizational readiness

is correct. And I try to Teams are also worried about ensuring responsible
AI. Only 30% of non-CEO senior executives say their
share what is going into organization is ready to adopt generative AI
responsibily.
the decision in case “You need people who can combine data, IT,

my team has doubts or and business process understanding,” says Matti

Lehmus, President and CEO, Neste, referring to

some different ideas.” AI adoption. “Then you can leverage your value
opportunities.” Without more pointed organization-
wide decisions about AI—and greater matching of
CEO goals with workforce characteristics and
Roberto Tomasi
planning—too many enterprises will likely miss their
Autostrade per l’Italia window of opportunity to lead in the AI transition.

69% 29%
CEOs Execs

I see broad benefits My organization already

from generative AI has in-house expertise to
across the organization adopt generative AI

Q: To what extent do you agree with the following statements?

(% reflects responses for “Agree” or “Strongly agree”).

28 29
 The IBM IBV CEO Study has
consistently asked leaders about their
most pressing priorities. Compared
with a year ago, productivity/
profitability has moved to the top of
the list, pushing customer experience
from the top position to third (see
Figure 5). Given the global economic
environment, the focus on the
bottom-line may be unsurprising.

But the areas that most bedevil CEOs are those

that are more ambiguous. As a company chairman
notes, “The biggest feature of turbulent times is the
increase in uncertainty. A large part of growth
comes from dealing with that uncertainty.”

Part Three

What
decisions
matter most?
30 31
 Figure 5
What are the CEO’s most Figure 6

Shifting CEO priorities challenging issues? CEOs’ top challenges

over the next 3 years
While CEOs identify productivity and profitability
as their top priority, they pinpoint sustainability
and cybersecurity as their biggest challenges
(see Figure 6). 42%
Environmental sustainability
Sustainability
2022 2023 Roughly 50% of CEOs and their executive teams now

Customer experience 58% 48% Productivity or profitability

have compensation tied to sustainability goals, a
significant jump from a year ago, when the figure was 32%
just 15%. “For us at smart, sustainability always has Cybersecurity/data privacy
Product and service innovation 50% 45% Tech modernization been and still is a hygiene factor: If you are not
credible there as an EV-only brand, customers will
Environmental sustainability 48% 44% Customer experience not pick you,” Adelmann of smart Europe tells us.
And in fact, our 2022 CEO study revealed that chief 27%
Customer relationships 45% 43% Cybersecurity and data privacy executives who align their sustainability strategy Tech modernization
with their digital transformation reported revenue

25%
More accurate forecasts 45% 39% Environmental sustainability growth of up to 41% higher than those who did not
align their efforts.7

Productivity or profitability 45% 38% Product and service innovation Yet while 95% of companies have operational ESG
Talent recruiting/retention

goals, only 10% have made significant progress

Improve marketing and sales 40% 30% Market share growth toward meeting them.8 This is a significant decision-
making disconnect. “I don’t think there’s any
24%
Diversity and inclusion
organization that will tell you sustainability isn’t

22%
important,” says CaixaBank’s Gortázar. “The first
thing to differentiate is who is just talking and who
really believes that. The CEO may be saying a whole Forecast accuracy
series of things that look good, but the reality is that
often it is not embedded in the operation.” 21%
Ecosystems and partnerships

Q: Which of the following represent your organization’s

highest priorities over the next 3 years?
Q: Which of the following represent your organization’s greatest
challenges over the next 3 years?

32 33
“Decision-making Trial Holdings Inc. is promoting efforts to reduce
energy use and decarbonization. For example, the
Cybersecurity Security-related data concerns
keep CEOs up at night
is not just taking company is utilizing abandoned educational Three out of four CEOs (76%) say that effective
facilities for its R&D site rather than renovating cybersecurity across their business ecosystems

data and trying to buildings. The company emphasizes the importance requires consistent standards and governance. While

54%
of utilizing existing assets as much as possible. That’s a tough issue for large national organizations

model accurately but

but an almost impossible one for global ones,
Complicating the issue is sustainability’s expanding who must deal with different sets of standards and
of government/regulators
definition and organizational uncertainty around
spending time also on
regulations. Generative AI and quantum computing
are demanding more data
appropriate metrics. There’s also doubt about the (due to the need for quantum encryption) are poised
transparency and
trustworthiness of sustainability reporting. Only
the uncertainties and
to dramatically complicate an already very
accountability around
about 45% of CEOs reported confidence in the challenging space.
data security and privacy
ability to report on ESG strategy and initiatives, as
working around those.” well as data security and privacy, with accuracy.9 CEOs need to rely increasingly on their Chief
In addition, public trust has fallen in what is being Data Officers (CDOs), as decisions about data
reported.10 Matti Lehmus, President and CEO, Neste, and cybersecurity multiply. Six out of 10 (61%)
explains the challenge: “Decision-making is not just CDOs said their organizational data is secure and
Matti Lehmus taking data and trying to model accurately but protected, but they also shared that they struggle
President and CEO, Neste spending time also on the uncertainties and working with data management issues such as reliability
around those. For instance, the sustainability lens (47%), regulatory barriers (37%), unclear data
continues to broaden into topics like biodiversity. ownership (36%), and data siloes/lack of data
We are just now as an industry and as society integration (33%)—further complicating protecting
developing good metrics—how to measure, what data and privacy.11
Only

55%
type of data to use.” Tracking and measurement in
“Cybersecurity is one of those races where you
the sustainability space is a huge issue, according to
can never win—you just need to know if you’re
the CEOs we interviewed.
ready,” says CEO and President Javier Tamargo of of CEOs are confident
Canadian transportation firm 407 ETR. “We’re their organization has the
transforming our systems and migrating to the cloud ability to accurately and
because we recognize that as a safer environment.” completely report the
information stakeholders
demand around data
security and privacy

34 35
 Case study

Identifying a single
“I don’t think there’s any source of truth
organization that will tell you
sustainability isn’t important.
Melbourne Water is a government authority that protects and
manages water resources for the City of Melbourne, in Victoria,
Australia. “We treat about 90% of Melbourne’s sewage at our Eastern
The first thing to differentiate is and Western Treatment Plants,” says Daniel Bradshaw, Energy
Accountant at Melbourne Water. “Energy use is a primary expense

who is just talking and who really across the organization and accounts for about half the Victorian water
sector’s total carbon emissions.”

believes that. The CEO may be Challenge: Melbourne Water pledged to help lead the state’s water sector in climate
change mitigation. To reduce energy consumption and leverage

saying a whole series of things renewable alternatives would require a detailed tracking tool that the
organization did not possess. “We were using a legacy data management

that look good, but the reality is system that was cumbersome, limited, and difficult to use,” Bradshaw
recalls. The system relied on manual data entry and file uploads, which
often led to poor data quality. “A single misplaced comma could send the
that often it is not embedded in entire data set askew,” Bradshaw says. What’s more, users needed
specific knowledge of the system’s functionality, limiting who could

the operation.” access and analyze the data.

Solution: Melbourne Water decided to create a single source of truth

for its energy use, using cloud-based data and analytics software.

Gonzalo Gortázar Results: Teams can now generate sustainability reports in hours, versus days,
CEO, CaixaBank and analyze historical activity to permit the targeted replacement of
many non-renewable energy sources in its treatment and transfer
system with renewable resources such as solar panels, hydroelectric
power stations, and sewage gas combustion engines. Melbourne Water
can also track the performance of its renewable projects.

36 37
 Leading an organization has never been
harder than it is right now—and in all
likelihood, the challenges will only
escalate from here. CEOs need to
expect and embrace uncertainty, and
they also need to embrace new tools
that enhance outcomes.

The emergence of generative AI as a business and

decision-making tool is a defining moment. The
choice? To be a passive consumer or a wise AI value
creator. CEOs need to define a clear plan and charge
ahead, have the right capabilities in place, and
handle disruptions of all shapes and sizes without
slowing down or losing focus on strategic vision.
Leaders can’t be distracted by shiny objects, lured
down dark alleys, tempted by easy fixes, or lulled
into complacency.

Part Four

Better
decisions for
a better world
38 39
“Sometimes the CEO has to make a
Ready for what
decision without a clear view to the final is coming?
result. That’s when you say, ‘I think this is “Sometimes the CEO has to make a decision In what ways are CEOs waiting to respond, and how

the direction. I don’t know how to achieve

without a clear view,” says Autostrade per l’Italia do CEOs turn those impulses and areas forward: to
CEO Roberto Tomasi. “You can’t be certain about prepare, pre-empt, think forward? Farsighted CEOs
the future, and sometimes you have to rely on your have generally built a vision of the world to come
this result, but I have to move. And I will ability to change direction midstream. We need that’s positive—and AI for good must be prioritized,
stakeholders to allow that we can make a mistake. to make that vision a reality. Use the action guide

try to better understand the situation.’ This is extremely important in situations where the
future is very uncertain.”
that follows to assess your own decision-making
and how best to take it to the next level.

With humans and AI working together, a new wave

Today, you can’t be certain about the of competitiveness has arrived.

future. Sometimes you have to make the

No single decision-making model will suffice for all
situations. CEOs must bet on the community of
talent within their team and invest in that team to
decision to head in one direction and rely enable higher performance. “Definitely, it helps to
have information to make better decisions,” says

on your ability to try to change direction President Carlos Zenteno of telecommunications

firm Claro Colombia. “For me, artificial intelligence

midstream if you need to.”

helps identify processes or patterns and
comparisons with others in the industry. There is
also a lot of intuition and a lot of human experience
that goes into decisions, but always supported by
numbers and analysis.”
Roberto Tomasi
CEO, Autostrade per l’Italia

40 41
Action guide
Here’s what top CEOs should prioritize to
elevate decision-making in the age of AI.

CEO – Get everyone smart on tech and data. Provide targeted training specifically designed for Technology – Change the enterprise mindset from “adding AI” to “starting with AI.” Initiate and
an AI world. Elevate those, such as your CDO, who have the expertise and insight to bring deepen conversations with your teams about the use of AI, to remove roadblocks to
leadership and data
together business strategy, technology strategy, and data strategy. progress and to ensure safety measures are in place. Promote responsible AI through
guardrails that align with the organization’s values and standards.
– Make “outcomes over activity” a mantra. Be prepared to terminate projects that are not
delivering the intended value, supporting strategic goals, or following ethical guidelines. – Check the dash. Use a digital dashboard to provide real-time integrated insights across
the organization.
– Make the data work for you, not vice versa. Use a broad range of planning approaches,
including forecasting and modeling, scenario-based planning, benchmarking, and data – Fix data shortcomings. In an era of generative AI, prioritizing data lineage and
mining. Recognize that no single decision-making model will suffice for all situations. provenance, customizable proprietary data, and data security is crucial.

– Identify AI use cases that align with your organization’s principles, broader technical
guidelines, and architecture. Prioritize applications where AI can boost competitiveness,
Metrics and – Eliminate layers between the data source and the decision maker while prioritizing innovation, and unique business value.
flexibility over control.
decision-making – Accelerate transition to zero-trust security across the enterprise and partner network
– Guide decision-making across the organization, setting a framework for how decisions to power secure interactions, workflows, and innovation. Ensure consistent standards
are made up and down the line, and how information about those decisions flows. and governance for effective cybersecurity, including the areas around generative AI and

– Rely on your Chief Data Officer (CDO) for decisions about data and cybersecurity, quantum computing.

including data management, data reliability, regulatory factors, data ownership, and data
integration.

– Ask the Chief Sustainability Officer (CSO) and Chief Financial Officer (CFO) to create a Ecosystem – Simplify, digitize, and partner to build a resilient enterprise. Leverage open innovation
balanced sustainability/profitability roadmap. and create new opportunities by connecting external and open data. Build a common
and partners
– Set the rules. Look for opportunities to define standards around sustainability, data platform using open hybrid technology that is consistent, scalable, and optimized for the

security and privacy, and all forms of AI. organization and partner ecosystem.

– Align targets with ecosystem partners, encouraging and reinforcing the use of consistent
metrics and incentivizing collective action. Increase security at every point in the
Talent and – Assess the potential impact of generative AI on your workforce. Act with a clear view of ecosystem by adopting zero trust security practices.
how to help your workforce with the disruption and inevitable transitions AI will bring.
workforces – Select key ecosystem partners and double down. Evaluate the strength of current and
– Implement “digital-first” solutions to increase efficiency, engage talent, and develop potential partners and invest in those that will make a difference and help form a winning
new skills. Empower those with complementary skillsets to co-develop AI and reimagine team going forward.
workflows.

– Know where your talent is coming from. Recognize potential skills shortages and align
top talent to areas most critical to competitive advantage.

42 43
 Research and Contributors
analysis methodology
The IBM Institute for Business Value, in cooperation with Oxford Economics, Many thanks to the many people whose invaluable insights,
interviewed 3,000 CEOs from over 30 countries and 24 industries as part of the expertise, and support helped make this report possible:
28th edition of the IBM C-suite Study series. These conversations, completed from
February through April 2023, focused on executives’ perspectives on leadership Cindy Anderson, Steve Ballou, Madhuri Banda, Santiago Baquero,

and business; their changing roles and responsibilities; and CEO decision-making Christian Bieck, Kristin Biron, Andrea Bocotti, Christopher Bontempo,

today, including key challenges and opportunities, their use of technology, data, Jürgen Braun, Jamie Cattell, Annette Celio, Bill Chamberlin, Marisa

and metrics, and their visions for the future. Insights were also drawn from Conway, Haynes Cooney, Wei Ding, Darko Djapic, Mandy Drouin, Brian

numerous client conversations including more than a dozen deep-dive interviews Goehring, Guilia Gotti, Prashant Gupta, Victoria Hyndman, Jayme

with CEOs conducted explicitly for this study from March through May 2023. Johnson, Vivek Kapur, Christian Kick, Peter Korsten, Stewart Lacey,
Rachel Larkin, Kris Lawas, Salima Lin, Jesus Mantas, Anthony Marshall,
Respondents in our study represent the most senior executives in their Kathy Martin, Michelle Mattelson, Dave McCann, Antoine Michaux,
organizations: CEOs, Public Leaders, General Managers, and Managing Directors. David Miller, Rakesh Mistry, Yu Muraoka, Hebatallah Nashaat, Masahiro
The IBM IBV designed data collection by geographic location and industry, with Numata, Florencia Paradeda, Talita Cristina Paro Fabene, Daritza Perez,
representatives across organizations of various sizes as defined by annual Roger Premo, Kristine Rodriguez, Wendy Roth, Thiago Sartori, Seiji Sera,
revenue or, in the case of public sector organizations, annual budgets. Christine Selph, Barbara Sheedy, Sherihan Sherif, Yan Xia Shi, Juha
Sipola, Ai Jun Sun, Noriko Suzuki, Yuta Torisu, Vanessa van de Vliet,
For this study, IBM IBV used a series of analytical methods to identify a group
Li Wang, Andrew Womack, and Valquiria Zuntini.
of respondents whose performance across a variety of factors in our 2023 IBM
CEO study survey instrument differentiated them from the overall respondent
population. Those in this group expressed strong confidence in both their existing
digital infrastructure and the breadth of their decision-making capability.

IBM IBV, in cooperation with Oxford Economics, also conducted a survey of 200
CEOs in the United States on their response to generative AI. These additional
responses were collected in April through May 2023.

44 45
 Notes and sources

1 All data cited in this report is from the IBM Institute for Business Value IBM Institute for The right partner
2023 CEO Study dataset, unless otherwise noted. Business Value for a changing world
2 Granger, John, Jesus Mantas, and Salima Lin. “Seven bets.” IBM Institute for
Business Value. May 2023. [Link] At IBM, we collaborate with our clients,
For two decades, the IBM Institute for Business
institute-business-value/en-us/report/seven-bets bringing together business insight,
Value has served as the thought leadership think
tank for IBM. What inspires us is producing advanced research, and technology to
3 “Hydra.” Britannica. Accessed May 26, 2023. [Link]
research-backed, technology-informed strategic give them a distinct advantage in today’s
topic/Hydra-Greek-mythology
insights that help leaders make smarter business rapidly changing environment.
4 Dorsey, Dustin. “What are the benefits and concerns of artificial intelligence?” decisions. 
abc7 News. April 4, 2023. [Link]
chat-gpt-open-ai-benefits/13059435/ From our unique position at the intersection of
business, technology, and society, we survey,
5 Granger, John, Jesus Mantas, and Salima Lin. “Seven bets.” IBM Institute for interview, and engage with thousands of
Business Value. May 2023. [Link] executives, consumers, and experts each year,
institute-business-value/en-us/report/seven-bets synthesizing their perspectives into credible,
inspiring, and actionable insights. 
6 DeRose, Adam. “These companies have banned or limited ChatGPT at work.”
HR Brew. May 19, 2023. [Link] To stay connected and informed, sign up
these-companies-have-banned-chatgpt-in-the-office to receive IBM IBV’s email newsletter at
[Link]/ibv. You can also follow
7 The 2022 CEO Study. “Own your impact: Practical pathways to
@IBMIBV on Twitter or find us on LinkedIn
transformational sustainability.” IBM Institute for Business Value. May 2022.
at [Link]
[Link]
en-us/c-suite-study

8 Biswas, Arun, Elisabeth Goos, and Jacob Dencik, PhD. “The ESG data
conundrum.” IBM Institute for Business Value. May 2023. [Link]
esg-data-conundrum

9 Ibid.

10 Ibid.

11 “Turning data into value: How top Chief Data Officers deliver outsize
results while spending less.” IBM Institute for Business Value. April 2023.
[Link]

46 47
 © Copyright IBM Corporation 2023

IBM Corporation
New Orchard Road
Armonk, NY 10504

Produced in the United States of America | June 2023

IBM, the IBM logo, and [Link] are trademarks of International Business
Machines Corp., registered in many jurisdictions worldwide. Other product and
service names might be trademarks of IBM or other companies. A current list of
IBM trademarks is available on the web at “Copyright and trademark information”
at: [Link]/legal/[Link].

This document is current as of the initial date of publication and may be changed
by IBM at any time. Not all offerings are available in every country in which
IBM operates.

THE INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT ANY

WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES
OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY
WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM products are
warranted according to the terms and conditions of the agreements under
which they are provided.

This report is intended for general guidance only. It is not intended to be a

substitute for detailed research or the exercise of professional judgment. IBM
shall not be responsible for any loss whatsoever sustained by any organization
or person who relies on this publication.

The data used in this report may be derived from third-party sources and IBM
does not independently verify, validate or audit such data. The results from the
use of such data are provided on an “as is” basis and IBM makes no
representations or warranties, express or implied.

1V2XKXYJ-USEN-04
48