Can AI Save Everyone's Bacon?

As 2024 rolls on, the enterprise markets for telecom and cloud services are a study in mixed messaging. On one hand, both telecom service providers and public cloud hyperscalers report that sales cycles are longer and more complicated. But both also claim strong customer interest in artificial intelligence (AI), particularly generative AI (GenAI), which creates code, text, or multimedia output from natural language input.

When it comes to actually investing in AI, however, hyperscalers have put their money where their mouth is, apparently funneling money into AI infrastructure while cutting workforce and other costs elsewhere. Clearly, for the hyperscalers, capital expenditure (capex) on AI infrastructure could edge out other investments this year.

Meanwhile, telcos are lagging in spending, though claiming interest in using AI for operations (AIOps), including network automation, customer service, predictive maintenance, improved security, and mobile network augmentation.

What is the financial situation for both camps? And what does that mean for potential capex output this year? Looking back, hyperscalers showed significant growth in 2023, while telco earnings sloped downward, indicating ongoing market preoccupation with cloud computing versus relatively flat interest in telco business services. Can customer enthusiasm for AI change any of this? That appears to be an open question.

The Telco Situation

Meanwhile, it's clear that telecom providers worldwide have reached a decisive moment: They are faced with the failure of 5G to live up to expectations. They are confronted by enterprise customers grown tightfisted in the face of macroeconomic woes. They also face unprecedented competition from the major public cloud providers as the center of technological gravity shifts to the cloud. And they’re falling behind in AI. With all these challenges, the time has come to change or continue to lose out. The situation is clear when revenues of a handful of leading telcos are compared with those of the leading cloud providers, as shown in the chart below:

The Hyperscaler Situation

The leading public cloud service providers, while also feeling the macroeconomic pinch and lowering their capex relative to former years, are ramping up spending on AI infrastructure. Their goal is to offer customers, including leading AI startups, the means to leverage the public cloud platforms to train large language models (LLMs) and develop applications from them. To this end, the cloud hyperscalers have been trimming other budget items in order to support their AI goals. Microsoft, for instance, eliminated over 10,000 jobs in 2023. Amazon laid off 9,000 last year, and Google about 12,000.

Still, it’s clear that the hyperscalers continue to spend more as a group than the telcos, even as both camps are reducing overall capex. This is unlikely to change, as enterprises – and AI startups such as OpenAI and Anthropic – continue to rely on cloud resources to develop AI solutions.

In our latest CTP Quarterly Update report, we take a detailed look at telcos and cloud providers’ financials, particularly capex spending, with an eye to what this information says about enterprise market needs. And we consider how well prepared both camps are to shift their strategies to meet the growing demand for AI.

CTP subscribers can download the report here.

As ever, let us know what you think. Email us individually or at analyst@futuriom.com