Beyond the hype: Capturing the potential of AI and gen AI in tech, media, and telecom

| Report

The emergence of generative AI (gen AI) presents both a challenge and a significant opportunity for leaders looking to steer their organizations into the future. How big is the opportunity? McKinsey research estimates that gen AI could add to the economy between $2.6 trillion and $4.4 trillion annually while increasing the impact of all artificial intelligence by 15 to 40 percent. In the technology, media, and telecommunications (TMT) space, new gen AI use cases are expected to unleash between $380 billion and $690 billion in impact—$60 billion to $100 billion in telecommunications, $80 billion to $130 billion in media, and about $240 billion to $460 billion in high tech. In fact, it seems possible that within the next three years, anything not connected to AI will be considered obsolete or ineffective.

Some leaders are moving to seize the moment and implement gen AI in their organizations at scale, but others remain in the pilot stage, and some have yet to decide what to do. If companies are to remain competitive and relevant in the coming years, it is essential that executives understand the potential impact of gen AI and develop the strategies necessary to incorporate it into their operations. Such strategies would involve an AI-native transformation, focused on building and managing the adoption of gen AI. McKinsey has conducted extensive research into how to embed gen AI to ensure that the technology delivers meaningful value. We’ve also spent much of the past year working with clients to create and then implement gen AI road maps. That combination of research and hands-on experience has allowed us to identify more than 100 gen AI use cases in TMT across seven business domains.1

Our experience working with clients already indicates the potential for telcos to achieve significant impact with gen AI across all key functions. The largest share of total impact will likely be in customer care and sales, which together would account for approximately 70 percent of total impact; network operations, IT, and support functions would round out the rest. The technology already is showing meaningful impact in enhancing interactions between employees and customers: the personalization of products and campaigns, improvements in sales effectiveness, and a reduction in time to market can spark a potential revenue increase of 3 to 5 percent. Customer care interactions—where as much as 50 percent of activity could be automated—have potential for a 30 to 45 percent increase in productivity while improving the customer experience and customer satisfaction scores. On the labor side, up to 70 percent of repetitive work activities could be automated via gen AI to improve productivity. There is also potential for new efficiencies in knowledge search, validation, and synthesis, where some 60 percent of activity has the potential for automation. And gen AI tools could boost developer productivity by 20 to 45 percent.

These areas provide rich soil for use cases. More challenging will be to go from sketching a road map to building proofs of concept to scaling successfully and capturing impact. Years of experience in designing and implementing digital transformations have taught us a lot, but gen AI’s nature and speed of disruption are creating a new layer of uncertainty.

Becoming an AI-native organization at scale involves making the most of technology, data, and governance. Success follows when leaders embrace an operating model that leverages the strengths of both humans and machines; is rooted in agility, flexibility, and continuous learning; and is supported by strong data and analytics talent. Another condition of success is to invest in data quality and quantity, focusing on the data life cycle to ensure high-quality information for training the gen AI model. Building capabilities into the data architecture, such as vector databases and data pre- and post-processing pipelines, will enable the development of use cases. Talent, data, technology, governance—none of these can be an afterthought.

Successful implementations share a clear vision and decisive approach. We advise that financial plans maintain or increase gen AI budgets over the next year. These budgets should include resources dedicated to gen AI for the shaping and crafting of bespoke solutions (for example, training large language models with telco-specific data, rather than implementing off-the-shelf ones) or partnerships with IT vendors to accelerate the timeline for implementation.

The AI journey has been shown to contain many challenges and learning opportunities, such as preparing and shifting an organization’s culture, finding data sets of significant size, and addressing the interpretability of the outputs provided by models. Leaders should expect such daunting challenges as a shortage of talent, lack of organizational commitment and prioritization (including among C-level executives), and difficulties in justifying ROI for certain business cases, all amid a changing regulatory and ethics landscape that creates further uncertainty. But daunting does not have to mean impossible. Developing a system of protocols and guardrails (such as building “moderation” models to check outputs for different risks and ensure users receive consistent responses) will be a crucial step toward mitigating the new risks introduced by gen AI. Another key will be change management—involving end users in the model development process and deeply embedding technology into their operations.

This collection presents the top insights on gen AI from McKinsey and its AI arm, QuantumBlack, providing a detailed examination of this technology’s transformative potential for organizations. It offers top management guidance on how to prepare for the implementation of gen AI and explores the implications of gen AI’s use by the TMT industries, especially telecommunications. The collection covers the essential requirements for deploying gen AI, including organizational readiness, data management, and technological considerations. It also emphasizes the importance of effectively managing risks associated with gen AI implementation. Furthermore, this compilation offers an overview of the future developments and advancements expected in the field of generative AI.

Gen AI will continue to evolve. New capabilities, such as the ability to analyze and comprehend images or audio, and an expanding ecosystem with marketplaces for GPT (generative pretrained transformers) are constantly emerging. For leaders, the stakes are high. But so are the opportunities. The next move from TMT players will define how they move from isolated cases to implementations at scale, from hype to impact.

For the full version of this report, download the PDF.

Explore a career with us