Recent market moves have exposed what Betashares described as a ‘growing contradiction’ in the AI investment narrative, with investors simultaneously questioning hyperscalers’ vast infrastructure spending while also pricing in the risk that AI could upend major software business models.
According to Betashares investment strategist Hugh Lam, those two concerns have become increasingly difficult to reconcile as rapid AI adoption accelerates across markets.
“There’s a growing contradiction in the AI narrative. On one hand, investors are worried about hyperscaler over-investment, and on the other they’re worried about software displacement. Those two fears are increasingly at odds with one another,” Lam said.
Lam said investor attention heading into 2026 had initially centred on whether a small group of mega-cap technology companies could justify enormous capital expenditure on data centres and AI infrastructure, but sentiment shifted during the first quarter as AI tools became more prominent and markets began focusing instead on which software companies could be displaced first.
“Heading into 2026, the market’s focus was whether a handful of companies commanding a large share of global equity markets could sustain enormous planned capital expenditure on data centres and AI infrastructure. But in the first quarter, as AI tools became more prominent, investors instead began questioning which software companies would be displaced first,” he said.
He argued that if AI adoption is moving quickly enough to threaten the world’s largest software names, it also implies the demand for compute power and infrastructure is substantial, making it harder to maintain the view that hyperscaler spending is excessive.
“If AI adoption is rapid and pervasive enough to threaten the business models of the world’s largest software companies, then the compute required to do that work is enormous and the technology is clearly valuable.
“That makes it harder to argue at the same time that hyperscaler spending is unjustified,” Lam said.
Rather than broad-based damage across US equities, Lam said the market had instead seen a sharp rotation beneath the surface, with hardware beneficiaries helping offset weakness among software names and keeping major indices relatively resilient.
“What we’ve seen is a sharp rotation under the surface rather than broad market damage. Broad US equity indices have been relatively resilient because hardware winners have helped offset software losers,” he said.
Lam said the recent software sell-off suggested investors were beginning to price in genuine AI disruption, but that same disruption also strengthened the investment case for data centre buildouts and cloud infrastructure.
He pointed to comments from Microsoft and Amazon Web Services indicating demand for cloud capacity continues to exceed supply.
“The software sell-off tells you investors are beginning to price in real disruption from AI, but that same disruption also strengthens the case that demand for compute, infrastructure and capacity is real. Microsoft has said demand for Azure capacity exceeds supply, and AWS has made a similar point.
“That suggests hyperscaler spending on data centres is looking increasingly justified, even if investors remain nervous about who ultimately earns the return on that capital,” he said.
Lam said the key question for markets was no longer whether AI infrastructure was necessary, but whether all players racing to build frontier models would generate adequate returns, warning that a winner-takes-most dynamic could leave some participants struggling to justify the scale of their investment.
He also said monetisation was already beginning to emerge, even as major providers continue subsidising usage to win market share, which Betashares sees as an early sign that businesses are embedding AI tools into day-to-day workflows.
“AI monetisation is already ramping, even while providers are still subsidising usage to build market share.
“That is an important signal, because it suggests businesses are adopting these tools quickly and embedding them into workflows.
“As reliance on AI tools deepens and switching costs rise, pricing power is likely to increase. That is where today’s subsidised adoption can start to turn into meaningful monetisation,” Lam said.
Betashares said its Nasdaq 100 ETF (NDQ) which holds 100 of the largest non-financial companies listed on the Nasdaq, includes a diversified portfolio of companies at the forefront of AI innovation and investment.
The fund has attracted nearly $200 million in net flows so far this year, has nearly $7.4 billion in funds under management, and has returned 18.32 per cent per annum since inception in 2015.
