Tech’s AI dreams spook investors – are they right to be fearful?

Big tech’s $660 billion spending surge sparks AI bubble worries among investors – but they’re short-sighted worries, asserts the CEO of one of the world’s largest independent financial advisory organizations.

The analysis from deVere Group’s Nigel Green comes as tech stocks sold off heavily after the major companies in the sector revealed plans to spend $660bn this year on artificial intelligence.

Amazon, Google, Microsoft and Meta are planning a surge in capital expenditure on data centres and specialist chips that pushes combined outlays well beyond last year’s $410bn, a jump of around 60% and a sum that rivals the output of mid-sized economies.

Amazon has gone furthest, warning that capex will hit $200bn this year alone, $50bn more than markets expected, eclipsing already vast commitments from Google and Microsoft.

Nigel Green says: “The scale has unsettled shareholders despite strong cloud revenues, wiped hundreds of billions from market values, and revived familiar questions about whether the AI arms race is drifting from strategic investment into excess without the genuine performance.

“While these concerns are understandable, and I believe the AI reckoning on profitability has already begun, it could be argued that these fears among AI investors are short-sighted.”

He continues: “The size of the number is what unnerves people, but the framing is wrong. This is not capital being sunk into a single product that has to quickly justify itself.

“What’s being built is a foundational layer that underpins everything that these companies do, and will do in the future.”

The deVere CEO argues that the market reaction ignores how this type of investment actually works.

“Much of the spending is concentrated upfront, yet it’s going into long-lived infrastructure. The accounting impact is spread over many years, even if the cash commitment happens immediately.”

He says expectations of a neat, standalone payback also miss the point.

“AI doesn’t need to show up as a separate revenue stream to generate returns. Its value appears through stronger customer retention, greater pricing leverage and lower churn across existing platforms.

“Even marginal improvements at that scale translate into substantial, recurring gains.”

The cloud businesses are where the economics become clearer.

As AI workloads mature, they lock customers in and command higher-value contracts. Over time, this supports stronger margins rather than eroding them. These cloud platforms already generate exceptional profitability, and advanced AI deepens that advantage.

Nigel Green adds that part of the spending is defensive by necessity. “Maintaining relevance requires scale. The market may dislike the competitive escalation, but for the companies involved, standing still is not an option.”

He concludes: “Volatility around these investments reflects uncertainty about timing, not a collapse in the underlying logic. Similar doubts surrounded earlier waves of infrastructure spending, which later proved foundational.

“This phase of AI investment is, I believe, laying groundwork that is likely to support earnings power for many years to come.”

“Under investing by these tech titans is perhaps the bigger risk.”