Despite mounting skepticism and rising failure rates, gen AI investment continues to skyrocket
Spending on generative AI (gen AI) worldwide is set to soar to $644 billion in 2025, representing a 76.4% increase from 2024, according to a new forecast by research firm Gartner. This explosive growth comes even as enthusiasm for gen AI’s capabilities cools amid high failure rates in early proof-of-concept (POC) projects and widespread dissatisfaction with current results.
“Expectations for [gen]AI’s capabilities are declining due to high failure rates in initial proof-of-concept work and dissatisfaction with current GenAI results,” said John-David Lovelock, VP analyst at Gartner. “Despite this, foundational model providers are investing billions annually to enhance GenAI models’ size, performance and reliability. This paradox will persist through 2025 and 2026.”
This paradox reveals a classic hype-cycle scenario: While expectations are taking a hit, investment continues based on long-term potential, competitive pressure and infrastructure build-out. It appears gen AI isn’t slowing down — it’s just entering a more pragmatic phase where enterprises are refining how and where they spend.
Gartner, for instance, reported that as gen AI moves deeper into enterprise operations, CIOs are expected to scale back internal development efforts and lean more heavily on off-the-shelf software solutions that integrate gen AI features. “Ambitious internal projects from 2024 will face scrutiny in 2025,” Lovelock said. “CIOs will reduce POC and self-development efforts, focusing instead on GenAI features from existing software providers.”
The report also revealed that hardware is dominating gen AI market growth, accounting for a staggering 80% of total investment. AI-enabled consumer devices such as smartphones, PCs and servers will lead the charge, as manufacturers increasingly embed AI capabilities by default.
“The market’s growth trajectory is heavily influenced by the increasing prevalence of AI-enabled devices, which are expected to comprise almost the entire consumer device market by 2028,” said Lovelock. Interestingly, he noted that this demand may be more manufacturer-driven than consumer-led. “[C]onsumers are not chasing these features,” he said. “As the manufacturers embed AI as a standard feature in consumer devices, consumers will be forced to purchase them.”
Here’s a breakdown of projected 2025 GenAI spending by category (millions of U.S. dollars):
|
2024 Spending |
2024 Growth (%) |
2025 Spending |
2025 Growth (%) |
Services | 10,569 | 177.0 | 27,760 | 162.6 |
Software | 19,164 | 255.1 | 37,157 | 93.9 |
Devices | 199,595 | 845.5 | 398,323 | 99.5 |
Servers | 135,636 | 154.7 | 180,620 | 33.1 |
Overall GenAI | 364,964 | 336.7 | 643,860 | 76.4 |
Source: Gartner (March 2025)
Telcos, too, remain committed to AI/gen AI
While concerns may be growing around gen AI’s true capabilities and impact, KT’s Chief Responsible AI Officer Soonmin Bae and Rakuten Symphony’s Chief Marketing Officer Geoff Hollingworth took the stage at last month’s Mobile World Congress Barcelona to directly address why, exactly, telcos should remain optimistic about how AI will transform their business.
One key point articulated by Hollingworth is that AI itself will solve the AI talent problem for telcos. “Telecom has been a laggard in technology innovation,” he admitted, but added that this has only really been for one reason. “Technology innovation is a software business; it’s a software domain. We have never been able to attract, retain and develop the kind of software engineers that the technology leaders have done outside ourselves… Our biggest opportunity is to use AI to become a leader in the next software wave, where we are all empowered because of our domain knowledge and we’re not limited now because of our lack with coding skills.”
Further, a month prior, Nvidia published a survey demonstrating that telcos are leaning into AI and gen AI quite aggressively, going against the industry’s reputation for being slower to adopt new technologies. In fact, Nvidia’s Global Head of Business Development for Telco Chris Penrose told RCR Wireless News that telcos, by many accounts, appear to be leading in the adoption of these technologies.
He said: “And part of the reason why is the opportunity. These are large companies with lots of legacy systems and lots of employees and lots of places where AI — and generative AI in particular — can really help transform the cost structure, transform the customer experience, make their employees more productive and ultimately, drop dollars to the bottom line and help them make more revenue… telcos have leaned into AI more aggressively … [and] we are starting to see some real results behind that.”