Massive AI capex not a bubble, but software & IT business models at risk: Arvind Sanger

Massive AI investments by US tech giants are unlikely to be a bubble waiting to burst, but the ripple effects could fundamentally disrupt traditional software and IT services models — including Indian IT companies, says Arvind Sanger, Managing Partner at Geosphere Capital.

Speaking to ET Now, Sanger said while headline narratives around US policy and investment promises continue, the bigger structural shift lies in how artificial intelligence is reshaping global business models.

No surprise in ‘America First’ pitch

Commenting on recent US policy rhetoric, Sanger noted that there were no major surprises in President Donald Trump’s remarks around tariffs and investment flows.

“Every president spins data to make it look good,” he said, adding that large investment promises from countries such as Japan, Korea and India need to be evaluated carefully over time.

Markets, he suggested, appear largely unsurprised by the messaging.

Is AI capex a bubble? “Very unlikely”

With AI-related capital expenditure from the so-called “Magnificent Seven” running into hundreds of billions of dollars, concerns are mounting about overinvestment.
Sanger disagrees.He argues that companies face a binary choice: Be the disruptor investing heavily in AI or become the disrupted by underinvesting.

“If AI adoption continues at the pace we’re seeing, spending may fluctuate year to year, but a collapse from hundreds of billions to negligible levels is very unlikely,” he said.

In his view, AI represents a structural wave of value creation rather than a speculative bubble.

The real risk: Business model disruption

The bigger concern lies not in hardware or AI model builders, but in software, SaaS and IT services firms whose pricing models depend on manpower-heavy structures.

Sanger illustrated this with an example from consulting firms using AI agents.

“If clients realise AI is doing a large part of the work, they will question why they are paying traditional billing rates,” he noted.

This creates pressure on:

• SaaS subscription models
• Consulting and advisory billing
• IT services revenue linked to headcount

Indian IT at crossroads

For Indian IT companies, which historically relied on large workforces and offshore delivery models, the AI shift could be transformative.

The key risks include:

• Fewer people required for similar output
• Revenue deflation if clients demand lower pricing
• Margin compression if business models fail to evolve

While some firms may successfully adapt and reposition themselves as AI integrators or platform players, Sanger believes the group as a whole faces structural challenges.

Valuations may need a reset

Recent derating in global software and IT services stocks reflects growing investor concerns.

Sanger suggested that some of the selloff may be justified, particularly if long-term “terminal values” of these businesses shrink.

“If the workforce declines structurally over five to six years, can revenues and profit pools remain at current levels? That’s the big question,” he said.

He expects meaningful disruption to play out over a multi-year horizon rather than immediately.

Short-term gains, long-term uncertainty

In the near term, AI implementation revenues may offer some support to IT services firms.

However, over the longer term:

• Companies will need to redesign pricing models
• Efficiency gains could reduce overall billing
• Profit pools across the sector may shrink

The AI wave may create enormous value — but not necessarily for everyone.

AI spending is not a bubble in Sanger’s view. But the implications for traditional software and IT services companies — especially manpower-heavy models — are profound.

Investors may need to differentiate sharply between companies that are building the AI future and those at risk of being disrupted by it.