AI Bust Expected Within 12 Months But Rebound Will Be Bigger Than Internet Era, Says NTT Data CEO

2 Sources

Share

NTT Data Inc. CEO Abhijit Dubey warns that the AI boom is headed for a bust as enterprise demand lags behind the capital rush into AI infrastructure, creating data center overcapacity. However, he predicts the correction will be short-lived, with a bounce back much bigger than the Internet era and compressed into a shorter timeframe.

NTT Data CEO Warns of Imminent AI Bust Followed by Historic Rebound

Abhijit Dubey, CEO of NTT Data Inc., has issued a stark warning that the AI boom is heading toward a bust within the next 12 months, but he expects the subsequent rebound to dwarf the Internet era's recovery. Speaking to the Economic Times, Dubey explained that enterprise demand for AI is failing to keep pace with the massive capital rush into AI infrastructure investments, leading to data center overcapacity across the industry . Despite this looming correction, the rare India-born executive to helm a Japanese multinational remains bullish on AI's long-term prospects. "The bounce back in AI will be much bigger than the Internet era and much shorter in time," Dubey said .

Source: ET

Source: ET

Infrastructure Spending Outpaces Corporate Adoption of AI

The disconnect between infrastructure spending and corporate adoption of AI has created what Dubey describes as a supply-demand imbalance. NTT Data Inc., which controls approximately 30% of India's data centers market, is investing $1.5 billion in India through FY27, part of an $11-billion global investment plan . This comes as hyperscalers like Google, Microsoft, Meta, and AWS pour billions into the region. Google recently committed $15 billion to build a 1 gigawatt AI hub in Andhra Pradesh, while Microsoft pledged $3 billion in India . In an interview with the Reuters Global Markets Forum, Dubey noted that with demand for compute still running ahead of supply, "supply chains are almost spoken for" over the next two to three years, with pricing power tilting toward chipmakers and hyperscalers

2

.

Why Enterprise Demand Lags Behind AI Infrastructure Investments

Dubey identified several factors preventing organizations from fully embracing AI despite massive infrastructure buildouts. Companies must overhaul 50-60 years of legacy systems, upgrade networks, fix data pipelines, and spend "as much on change management as on technology" to achieve meaningful outcomes . A PwC survey released in November supports this assessment, revealing that daily use of GenAI remains "significantly lower" than widely touted by executives, even as the technology reshapes expectations

2

. Dubey expects adoption to accelerate well before five years, provided enterprises focus on building necessary foundations.

AI Reshaping Labor Markets Despite Slow Enterprise Uptake

The impact of AI reshaping labor markets is already visible, though adoption patterns remain uneven. Dubey acknowledged that NTT Data Inc. has begun rethinking recruitment strategies as AI transforms work. "There will clearly be an impact... Over a five- to 25-year horizon, there will likely be dislocation," he told Reuters

2

. However, he emphasized that NTT will not cut roles and has grown its headcount in India, hiring more this year than in the past . The AI skills gap is widening, with workers possessing AI capabilities commanding an average wage premium of 56%—more than double last year's figure, according to PwC

2

. About half of non-managers report access to training resources, compared with roughly three-quarters of senior executives.

Source: Market Screener

Source: Market Screener

NTT Data's Strategy: Enterprise Inferencing Over Frontier AI Model Training

While hyperscalers invest billions in frontier AI model training, NTT Data Inc. is deliberately avoiding that segment. Instead, the $30-billion IT services and data center arm of the $100-billion NTT Group focuses on enterprise inferencing, which Dubey calls a "less risky" and more sustainable business with better long-term economics . "We've always stayed very true to enterprise inferencing... Even if there is a short-term overbuild, (in the) long term, it absolutely will be used up," he explained. Dubey noted that concerns about overinvestment in data center capacity are overstated because while chips have a 3-4 year life cycle, data centers remain productive for 15-20 years .

Massive Expansion Opportunity Beyond Traditional IT Services

Dubey sees AI as an opportunity for market share gain rather than a threat to the $283-billion Indian IT industry. While global IT spend stands at $5 trillion, non-IT costs for businesses total $35-40 trillion . Even a 1-2% productivity improvement in this non-IT spend would unlock an addressable market far larger than today's services industry, representing what he calls a "massive expansion opportunity." Speaking to Reuters, Dubey emphasized that "there is absolutely no doubt that in the medium- to long-term, AI is a massive secular trend"

2

. The AI market rebound, he predicts, will be both stronger and faster than previous technology cycles.

India's AI Ecosystem Faces Silicon Valley Challenges

On India's AI startup ecosystem, Dubey candidly admitted that competing with Silicon Valley remains difficult because capital, customers, and innovation are overly concentrated there. "Most of the AI innovation is centred in Silicon Valley. Access to capital and access to US customers is unparalleled... That has to be fixed in India," he said . Despite workforce restructuring concerns across the industry, NTT Data Inc. expects to retain its relevance and leadership in data centers even as it faces competition from tech giants. "Just putting in capital doesn't make you an operator overnight," Dubey noted .

Today's Top Stories

TheOutpost.ai

Your Daily Dose of Curated AI News

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Spend less time searching for the latest in AI and get straight to action.

© 2025 Triveous Technologies Private Limited
Instagram logo
LinkedIn logo