Technology, Media & Telecommunications Practice

How generative AI could

revitalize profitability for telcos
The new technology offers the sector a real opportunity to reverse its stagnant
fortunes. But seizing it will require embracing innovation and agility to an
unprecedented degree.

This article is a collaborative effort by Stephen Creasy, Ignacio Ferrero, Tomás Lajous, Víctor Trigo, and Benjamim
Vieira, representing views from McKinsey’s Technology, Media & Telecommunications Practice.

© Getty Images

February 2024

49
Amid the intense competition and cost-cutting enhancing customer revenue through improved
confronting telcos, early evidence suggests that customer life cycle management and decisively
generative AI (gen AI) could be the catalyst to reducing costs across all domains.
reignite growth after a decade of stagnation. But
will it be a groundbreaking differentiator or simply However, while nearly all of the 130 telcos we
table stakes? surveyed are doing something with gen AI, our
survey findings suggest a palpable sense of
Already the technology is on track to become caution and skepticism in the industry. More
a new norm in the industry. Most telco leaders than 85 percent of the executives surveyed
we surveyed1 say they are developing gen AI are cautious to attribute more than 20 percent
solutions that range from pilots to full-scale revenue or cost savings impact by domain,
deployments, and leading telcos such as AT&T, with the greatest enthusiasm for a radical
SK Telecom, and Vodafone have made much- transformation in customer service (Exhibit 1).
publicized early gen AI commitments and
launched trials. Some telcos around the world This blend of optimism and restraint highlights
have started to experience significant double- the critical juncture the industry faces. Seizing
digit percentage impact from this technology. the gen AI opportunity to differentiate services
One European telco recently increased and achieve sustainable growth will require
conversion rates for marketing campaigns by the hidebound industry to embrace innovation,
40 percent while reducing costs by using gen AI exploration, and agility at an unprecedented level
to personalize content. A Latin American telco and move from decoupled AI efforts to a holistic,
increased call center agent productivity by 25 AI-native telco.2
percent and improved the quality of its customer
experience by enhancing agent skills and The chance for telcos to make this change has
knowledge with gen-AI-driven recommendations. never been more accessible. The industry has
struggled these last ten-plus years to achieve the
Most impressive is that these telcos deployed the potential of “traditional” AI, given the complexity
models in just weeks—the first went live in two and legacy processes involved. In addition to
weeks, and the second in five. For an industry the significant impact gen AI can bring to bear
with a mixed track record for capitalizing on with entirely new use cases and applications,
new technologies and legacy systems that slow its ability to learn from vast amounts of diverse
innovation, these early results and deployment data and interact in near-human-like ways may
times illustrate the potentially transformative be the tipping point for accelerating broader AI
power of gen AI. programs and the building blocks that enable
them, fueling company-wide transformations.
These aren’t one-offs. Pretrained models that
can be fine-tuned in days for use cases are Furthermore, the imperative for such change has
readily available, enabling organizations to never been greater. Because gen AI democratizes
bring proofs-of-concept to life with minimal access to powerful capabilities, any telco—a
up-front investment, achieve impact out of the small operator or large incumbent—can reshape
gate, and scale their efforts. Our experience customer expectations and its organizational
working with clients indicates the potential for efficiency. In doing so, they can potentially narrow
telcos to achieve significant EBITDA impact previously unassailable competitive advantages
with gen AI. In some cases, estimates indicate and overturn long-standing barriers to growth.
returns on incremental margins increasing 3 to Those at the forefront of this movement stand to
4 percentage points in two years, and as much position themselves to regain growth faster and
as 8 to 10 percentage points in five years, by capture a more significant share of the nearly

1
The online survey was in the field from November 9, 2023, to December 6, 2023, and garnered responses from 130 telco operators in
North America, Latin America, Europe, Africa, Asia, and the Middle East.
2
See “The AI-native telco: Radical transformation to thrive in turbulent times,” McKinsey, Feb. 27, 2023.

How generative AI could revitalize profitability for telcos 50

Exhibit 1
A large
large majority
majorityof
of telcos
telcos have already cut costs with generative
costs with generative AI
AI use cases in
cases in customer service and
customer service and networks. networks.

Cost reduction attributed to generative AI

in different domains,¹ % 0 1–5 6–20 21–50 50+ Don’t know

2 1 1 2
6 2 3 1 3 6
8
7 6
13 15
15

32

35 30

40

87

39

50
47

34

16

Customer Network IT Marketing Support

service and sales functions

1
Gen AI CxO Survey 2023, n = 130, Q: What is the impact (% cost reduction) attributed to generative AI in the different domains? Percentages consider
answers only from respondents claiming to have achieved impact and to have at least some use cases in execution; figures may not sum to 100% because
of rounding.
Source: McKinsey analysis

McKinsey & Company

$100 billion in incremental value (Exhibit 2). Gen AI today in the telco industry
That is in addition to the $140 billion to $180 Gen AI represents the latest advance in AI, and
billion in productivity gains that gen AI will it may arguably be one of the most important.
create in the industry above what could be The technology’s ability to analyze more and
unlocked by traditional AI. different types of data such as code, images,
and text, and to create new content, enables
How can telco leaders use the technology new levels of personalization, performance,
to drive AI transformations and unlock new and customer engagement. With today’s
value? What challenges do they face? And capabilities, many use cases are already
what will it take to succeed? This article offers possible across network operations, customer
insights into these critical questions, drawing service, marketing and sales, IT, and support
extensively from our research, industry survey, functions.
and firsthand experience implementing these
technologies.

How generative AI could revitalize profitability for telcos 51

Exhibit 2
Generative AI
AI has
has the
the potential
potential to
to unlock
unlockvalue
valuebeyond
beyondthat
thatpreviously
previously
offered by advanced analytics and ‘traditional’
offered by advance analytics and ‘traditional’ AI.AI.

Generative AI’s potential impact on global telecommunications industry,¹ $ billion

140–180 450–680

60–100 310–500

250–400 ~35–70%
incremental
~15–40% economic
incremental impact
economic
impact

Advanced analytics, New Total All worker productivity Total AI

traditional machine learning generative use-case-driven enabled by generative AI, economic
and deep learning¹ AI use cases potential including in use cases potential

1
Updated use case estimates from “Notes from the AI frontier: Applications and value of deep learning,” McKinsey Global Institute, April 17, 2018.
Source: McKinsey Global Institute, The economic potential of generative AI, June 14, 2023

McKinsey & Company

These use cases can both enhance existing AI — In marketing and sales, where gen AI enables
capabilities (through the inclusion of new hyperpersonalization, deeper customer insights,
unstructured data sources) and provide new sources and faster content generation. A European telco
of value (through gen AI and in combination with is using the technology to identify new sales
traditional AI solutions) to deliver significant impact leads from customer calls, with its pilot project
across all key domains. Customer service and achieving a more than 10 percent conversion rate.
marketing and sales currently make up the largest The company can now also create personalized
share of total impact (Exhibit 3). messages and visual media to target individual
customer microsegments. To do this, the telco
Examples of such use cases based on early pilots feeds a gen AI model standard marketing
include the following: messages, customer data (including household
details, type of phone they use, and where they
— In customer service, where the technology can live), and cognitive biases (for example, whether
vastly improve customer experience, increase the customer would be more receptive to
agent productivity, and enable fully digital messaging that evokes scarcity, such as a limited-
interactions. A Latin American telco is enhancing time offer, or emphasizes authority, such as
its customer service AI chatbots to improve agent endorsements, awards, and industry experience).
support, a move it anticipates will reduce costs by
15 to 20 percent. The telco also is using gen AI to — In network operations, where gen AI can optimize
summarize voice and written client interactions technology configurations, enhance labor
in nearly a dozen use cases, with the expectation productivity, extract customer insights from
that it can reduce associated costs by up to 80 social media, and improve inventory and network
percent. planning and management through the ability to

How generative AI could revitalize profitability for telcos 52

Exhibit 3
Gen AI
Gen AI is
is expected to enable
expected to enableaalong
longlist
listof
ofuse
usecases
cases and
and deliver
deliver significant
significant
value to
value telcos,with
to telcos, with customer
customerservice
serviceand
andmarketing
marketingand
andsales
sales accounting
accounting
for the
for the largest
largestshare
shareofoftotal
totalimpact.
impact.

Distribution of impact by business domain¹ Share of total impact, % Share of surveyed business
leaders focused on domain, %
Customer Marketing Support
service and sales Network IT functions

~35 ~35 ~15 ~10 ~5

~85 ~45 ~62 ~55 ~10

Example use cases

Customer-facing Content generation, Network inventory Copilots for software Procurement

chatbots, call-routing hyperpersonalization, mapping, network development, optimization,
performance, agent copilots for store optimization via synthetic data workplace productivity,
copilots, bespoke personnel, customer customer sentiment generation, code internal knowledge
invoice creation sentiment analysis analysis, enabling migration, IT management, content
and synthesis self-healing via support chatbots generation, HR Q&A
customer sentiment
analysis on network
problems

1
The distribution of impact by business domain is based on our experience working with telco companies to deploy gen AI and includes impacts on both capital
expenditure and EBITDA.

McKinsey & Company

mine unstructured data. One large telco is resource constraints. Organizations are
using the technology to accelerate network applying gen AI to streamline the entire
mapping by analyzing and structuring data software life cycle, from documenting how
about network components, including a new product, feature, or service will be
specifications and maintenance information, perceived by end users to generating and
from supplier contracts. This will enable the scanning code for vulnerabilities before
telco to more accurately assess component launch. One McKinsey study found that
compatibility, maintenance requirements, software developers can complete coding
and more—an effort anticipated to improve tasks up to twice as fast with gen AI. 3
operational planning and optimize capital
productivity. — In support functions, where gen AI will
reduce the costs associated with back-
— In IT, where the technology can accelerate office operations and improve employee
software migrations and development. Gen AI productivity. A European telco uses the
offers telcos a path to reduce their mounting technology in a number of ways, including:
technical debt and enable capabilities shortening procurement analysis and
previously deferred because of time and negotiation strategy insights from weeks

3
“Unleashing developer productivity with generative AI,” McKinsey, June 27, 2023.

How generative AI could revitalize profitability for telcos 53

 to a few hours, reducing recruiting costs with encounter difficulties in justifying use cases—a
automated screening and recommendations, clear signal that much of the push has come from
improving employee productivity using internal the bottom, not the top, and that more work is
gen AI chatbots and copilots, and automating needed to elevate gen AI to a CEO-led priority.
most internal content generation. Combined, Moreover, finding appropriate talent and obtaining
the company anticipates these efforts will quality data remain significant challenges for telcos,
improve employee productivity by although confidence about solving these rose
30 percent. among surveyed leaders this year as compared
to last.
New sources of value may also emerge from turning
internal uses cases into new products for their Finally, the survey findings suggest that gen AI has
customers. For example, a customer care solution already begun to influence long-standing market
may be offered on demand to small business dynamics. While European telecom operators
customers seeking ways to improve their own call have traditionally lagged in AI and technology
center’s productivity and service. transformations, survey findings indicate that
they are pulling ahead of those in North America
in their adoption of gen AI, especially in areas
Telco leaders’ view ahead: Real change such as network operations (71 percent compared
and real challenges to 58 percent) and IT (67 percent compared to
In the wake of their initial successes, business 55 percent). This shift may be a result of greater
leaders we surveyed say they plan to maintain maturity in managing data privacy. Small and
or double their budgets for gen AI in the next large operators report similar views on where to
year and invest in more than 50 dedicated full- prioritize, focusing on customer service and IT in
time employees to pursue their gen AI ambitions similar measure, suggesting the possibility of new
effectively. More than half are already scaling up competitive pressures emerging for incumbents
use cases. (Exhibit 4).

Moreover, survey findings indicate that the

technology also had a knock-on effect across all AI The building blocks for a successful
initiatives. Compared to responses from McKinsey’s gen AI journey
2022 digital twin survey, we see a 30-percentage- In order to achieve the above-mentioned impact,
point increase in business leaders who want to organizations will need to move away from the
invest in and focus more on data and analytics. labyrinth of proofs-of-concept and scale the
technology. As with any digital or AI initiative, we
However, despite the magnitude of the opportunity find there are no shortcuts in doing this. The same
and the level of interest (and need), our survey core building blocks are necessary, namely (1) a
found few who follow the kind of holistic approach business-led roadmap, (2) the right talent, (3) an
required to succeed at scale. Only about one-third operating model for scale, (4) technology built
of telco leaders said they have a capability-building for speed and innovation, (5) quality data that is
plan for employees on gen AI or are investing in easily accessible and managed in a responsible
change management efforts—two core building and accountable way, and (6) change management
blocks for building a culture of innovation and the to ensure adoption and scaling. 4 These are
test-and-learn mindset. A similar share said gen AI fundamental pillars in effectively scaling use cases
has yet to be treated as an organizational priority, and capturing sustainable impact from gen AI in the
and that proponents of the technology often journey toward an AI-native telco.

4
See “Rewired to outcompete,” McKinsey Quarterly, June 20, 2023.

How generative AI could revitalize profitability for telcos 54

Exhibit 4
Small and large operators
operatorsarearefocused
focusedon
ongenerative
generativeAIAIuse
usecases
casesforfor
customer
service
customerandservice
IT in similar measure.
and IT in similar measure.

Focus on gen AI for customer

service and IT use cases,¹ % < $1 billion $1 billion–$5 billion $5 billion–$10 billion > $10 billion

93
85
79 79
75

60 58
57 57
53 53 54 51
54

37
29

11 13
7 7

Customer service Network IT Marketing and sales Support functions

1
GenAI CxO Survey 2023, n = 130, Q: Focus domains in gen AI—Select the dimensions that apply depending on how you define generative AI in your
organization. Percentages may not sum to 100% as this question contained multiple selections.
Source: McKinsey analysis

McKinsey & Company

However, while the same holistic approach is foundation model from scratch (what we refer to
required, gen AI’s unique capabilities—its ability as the “taker,” “shaper,” and “maker” approach,
to surface new insights from seemingly unrelated respectively).
data, its reliance on large language models from
third-party vendors, and its transformative impact Each is suitable for different use cases and has
on roles and work—present new challenges that its own costs, requiring leaders to develop not
will require greater agility and additional oversight. only a clear vision and strategy for which use
Next, we outline key differences and provide cases to pursue, but also how. One mistake we
recommendations on how telcos can best see some telcos making is building common gen
tackle them. AI solutions from scratch—a content generator or
call summarization solution for example—when
Strategy: Determine when to build, buy, or fine- there are nearly a dozen out-of-the-box options
tune solutions on the market today from gen AI start-ups or
As with any AI initiative, leaders will need to align SaaS vendors injecting gen AI capabilities into
on vision, value, and road map, assessing both existing solutions. Only one-third of surveyed
risks and opportunities, and communicating telco leaders say they buy products off the shelf,
guidelines for use across the organization. In suggesting that many telcos continue to embrace
building a road map, telco leaders will face a a do-it-yourself model. This move is likely to
choice: use a commercial, off-the-shelf solution slow innovation and distract talent from more
if one exists, fine-tune existing large language differentiating use cases, as it has in the past with
models with internal data, or build and train a new other technologies.

How generative AI could revitalize profitability for telcos 55

Instead, leaders should strongly consider Capturing the full potential will also require
partnering with gen AI solution providers and significant upskilling of existing staff—everyone
enterprise software vendors for solutions that from data scientists to business leaders—on gen
aren’t very complex or telco specific. This is AI, including the risks of uploading proprietary
particularly critical in instances where any delays data into third-party language models. Some
in implementation will put them at a disadvantage telcos are setting up internal certification and
against competitors already leveraging these university-led training programs to ensure their
services. The handful of solutions leaders can teams have the right skills and capabilities to
concentrate on shaping or making themselves innovate and execute with the technology. For
should enable them to differentiate their offerings instance, a large telco created a badging system
or address a strategic business priority, such as to identify gen-AI-ready employees who have
delivering the best service or network coverage, completed the company’s sessions on use, risk,
and drive sustained economic impact. To do and effective prompting techniques given by its
this, one CIO at a large telco is bringing together AI, legal, and risk experts. Following certification,
business leaders across all key domains to assess users participate in weekly discussion groups
hundreds of potential use cases and build a road to stay abreast of changes and discuss their
map for determining when to build, buy, or fine- successes and challenges. McKinsey research
tune models, and for prioritizing resources and has found that such efforts improve the quality of
building momentum with early successes. prompts.

Talent: Upskill and expand internal expertise to Operating model: Orchestrate efforts
innovate with gen AI enterprise-wide
The speed of innovation that is now possible with A significant portion of implemented gen AI
gen AI puts new pressure on telcos accustomed solutions can be adapted and reused in multiple
to outsourcing tech talent to build in-house use cases. A gen AI chatbot developed to
AI expertise. Consider the experiences of two improve agent productivity, for example, can be
telcos—one that continued offshoring and repurposed with additional fine-tuning or data
outsourcing tech talent and one that created to answer frequently asked questions by new
a dedicated AI team of ten data scientists and employees or provide IT support. An off-the-shelf
engineers. In the time the first telco took to draft content generation system for drafting sales
requirements for outsourcing gen AI use-case proposals may also streamline the development
development, the second built and deployed four of marketing and business plans.
gen AI solutions.
As a result, we’re beginning to see telcos adopt
While this new technology democratizes AI by more centralized decision making around gen
requiring fewer highly specialized data scientists AI development. This shift includes a greater
to build the models, it requires new skills, such as emphasis on adopting reusable services and self-
gen AI prompt engineering, which may sometimes service components, an evolution of key functions,
be a separate skill embedded within traditional such as risk, FinOps, and transformation offices,
roles. It also requires significantly more data to be more focused on gen AI, and the creation
engineers and subject matter experts who of “control towers” that can oversee all gen AI
understand what data to collect and how, and investments and development efforts. In practice,
who can oversee daily quality reviews as new that can mean, for example, prioritizing the
forms of data are generated and consumed by use-case pipeline, identifying opportunities for
these systems, including user queries, responses, reusability, setting key performance indicators to
and feedback. measure and track impact at the level of both use

How generative AI could revitalize profitability for telcos 56

case and enterprise, and managing suppliers and Once new models are deployed, telcos will need
risk. A European telco’s control tower evaluates to monitor model outputs daily to ensure quality
the effects of its gen AI transformation based on and accuracy do not waver as models learn and
three dimensions—financial impact, user adoption, adapt their responses based on user queries
and model performance—and aggregates the and feedback. Large language model operations
data in dashboards that enable the company’s (LLMOps) is an emerging practice that aims to
top executives to keep tabs on the organization’s streamline the daily management and monitoring
progress. Similarly, a Latin American telco uses of gen AI models. A key component of LLMOps is a
a control tower to consolidate and standardize dedicated operations team to oversee all deployed
supplier contracts, tracking key metrics such as gen AI models, continuously monitoring for issues
scope, duration, and renewal to compare providers and rapidly adapting solutions when needed, just
more easily, identify potential redundancies, and as a network operations team might do for network
reduce the manual work of digitizing content. performance. Organizations can start small
now and build capability in this area as the field
Technology: Create a blueprint for reusability, of LLMOps develops. For example, a European
innovation, and excellence telco started by assigning three data scientists to
Organizations will also need a technology blueprint monitor its handful of deployed models and plans
that enables reusability. For instance, the blueprint to expand the team as more models are deployed.
should include a framework for determining
which large language models to use and when Data: Capture everything, especially
(commercial or open-source models, for example, unstructured data, and ensure responsible use
or those that support hybrid workloads). And it One of gen AI’s superpowers is its ability to
should lay out how to scale a pilot—for example, uncover connections in seemingly unrelated
to extend a pilot that serves 100 call agents to datasets, which has implications for how
serve more than 10,000 agents with the same organizations choose to collect and measure data,
latency and cost profile. The blueprint should and how they manage it to ensure responsible use.
also have a framework for determining which
gen AI capabilities can be turned into ready-to- Data collection: Telcos will need to think more
use modules to be plugged into different use broadly about data collection, mapping more
cases. One large telco, for example, has begun data, setting up pipelines for unstructured data,
to identify and develop components designed and creating synthetic data to evaluate outputs.
to fetch product data from a large dataset and A US telco, for instance, has reached beyond
generate content from it that could be reused its traditional datasets in its work to develop a
by data science teams across domains such as customer service agent copilot that will reduce
customer service, network operations, and sales average resolution times by 40 percent across
and marketing. more than one million annual chats. As part of
their work, the company’s data scientists gather
With new gen AI research and capabilities institutional knowledge from agent emails and
being announced weekly and sometimes daily, interactions to enable the chatbot to learn from
technology teams will also need a dedicated real situations and challenges, and offer detailed
gen AI innovation lab to keep abreast of industry descriptions of how to resolve specific issues.
changes and test emerging solutions. For example, The team also creates synthetic data using a
one large telco’s chief data and analytics officer large language model to create sample customer
recruited PhD graduates from universities to staff questions and answers, with agents reviewing the
a gen AI innovation lab and build bespoke solutions outputs for accuracy.
ahead of the market to gain a competitive edge.

How generative AI could revitalize profitability for telcos 57

Responsible use: These types of more For example, another European telco saw
sophisticated data strategies and tactics come firsthand the importance of change management
with new regulatory, intellectual property, and and upskilling when it created a gen-AI-driven
data privacy concerns. Risks abound in this knowledge “expert” that helped agents get
new era, particularly with customer insights, answers to customer questions more quickly.
recommendations, and network optimizations The initial pilot, which didn’t include any process
being analyzed and generated by third-party changes or employee education, realized just
large language models and open-source a 5 percent improvement in productivity. As
environments. To address the novel risks, telcos the organization prepared to scale the solution,
need to expand their data governance programs leaders dedicated 90 percent of the budget
to address unstructured data. For example, one to agent training and change management
multinational telco hardware provider created processes, which facilitated the adoption of the
a robust data access process to validate what solution and resulted in more than 30 percent
data can be used in gen AI use cases. Data productivity improvement. The telco also used
owners along with legal and security experts gen AI to create upskilling programs and provide
work together to validate each use case based agents with personalized recommendations for
on several criteria including the criticality of the improvement once the solution was rolled out.
data to the business (data that is deemed of high
importance cannot be input into commercial large
language models); the end users (some users
cannot access certain data assets); and the risks Even though so many companies have already
if the gen AI solution gives an incorrect answer. achieved real cost savings and revenue
The team manages the process in an agile manner improvements with gen AI, these are still the early
using a simple Microsoft Power App to manage days of the technology. In the next five years,
and automate the workflow across teams, and emerging capabilities—including significant
conducts monthly forums to review the process improvements in natural language understanding,
and develop improvements. The organization has advances in human-like reasoning across
reviewed more than 200 use cases, rejecting a multiple topics, and availability of real-time
number due to intellectual property and other solutions with increased accuracy and fewer
risks, to ensure responsible use for the company. hallucinations—should unlock even more exciting
opportunities beyond the basic improvements
Change management: Ensure adoption and seen today.
scaling are CEO-led
Every role, including everyone from network Combined, these gen AI capabilities will enable
technicians to HR professionals, will be impacted telcos to redefine industry standards and set
by gen AI, making vital the need for leaders to themselves apart in the market. For example,
begin preparing their employees now to capture network operations could be enhanced and
the full value of this transformative technology. quality standards radically recast with AI copilots
With many employees already using the that evaluate images from technicians, provide
technology in their personal lives, organizations accurate recommendations for remedies, and
will need to consider how to help them learn to automatically initiate interventions or work
apply the technology in a professional context, orders. In sales, cognitive copilots could conduct
upskilling and reskilling staff at scale. Such sentiment analysis on customer calls in real
work can be made easier using gen AI, for time and guide sales representatives on how
example to develop and deliver customized and best to respond, profoundly altering sales
adaptive training programs, and even to onboard strategies, customer engagement, and overall
employees. sales outcomes. Customer service channels

How generative AI could revitalize profitability for telcos 58

using cognitive chatbots could seamlessly closer to becoming a reality. With it comes the
answer complex queries in real time while taking opportunity for telcos to reverse their recent
into account privacy and fairness concerns, stagnant fortunes and usher in a new era of
thereby revolutionizing efficiency while offering growth and innovation. The journey will not be
customers a human-like experience. Across the easy, however. To answer the call of gen AI, telcos
enterprise, greater efficiency and productivity will need to quickly adopt a culture of innovation
could emerge, as domain-specific solutions and experimentation enabled by the core
endowed with an organization’s institutional building blocks shared in this article, one they
knowledge power an unprecedented wave of have previously struggled to build and maintain.
automation and AI-driven decision making. With the technology moving so rapidly, those
operators that embrace it now are likeliest to
The sudden rise of gen AI has brought the create a significant lead that will be difficult for
dream of the AI-native telco significantly others to follow.

Stephen Creasy is a partner in McKinsey’s Copenhagen office, Ignacio Ferrero is a partner in the Miami office,
Tomás Lajous is a senior partner in the New York City office, Víctor Trigo is an associate partner in the Madrid
office, and Benjamim Vieira is a senior partner in the Lisbon office.

The authors wish to thank Joshan Cherian Abraham, Eric Buesing, Michael Chui, Guilherme Cruz, Sebastian Cubela,
Andrea Fariña, Roger Roberts, and Kayvaun Rowshankish for their contributions to this article.

Copyright © 2024 McKinsey & Company. All rights reserved.

How generative AI could revitalize profitability for telcos 59