Interview: ‘Big Tech’s AI capex is growing much faster than revenue,’ warns AvaTrade’s Kate Leaman

Big Tech is pouring huge amounts of money into AI right now, and we’re already seeing a mix of quick wins and longer-term bets play out.

Take Microsoft, Meta, and Amazon, their spending on AI infrastructure is skyrocketing.

Some of that is paying off quickly, especially in cloud services and AI-powered advertising. But overall, they’re still investing a lot more than they’re making directly from AI at this stage.

Alphabet is another example. It has a record $155 billion backlog in its AI cloud business, which shows strong demand from companies.

The challenge? Turning that demand into real profits, not just future potential.

Apple is taking a slower, more privacy-focused AI approach. While that fits its brand, it may fall behind if it doesn’t move faster and build a broader AI ecosystem.

Meanwhile, Meta is benefiting from AI-driven advertising improvements, but tighter global privacy regulations could create some headwinds.

In an exclusive interview with Invezz, Kate Leaman, chief market analyst at AvaTrade, decoded how Big Tech’s investments in AI are currently unfolding in terms of operational wins and future payoffs.

Excerpts:

Invezz: Big Tech (Microsoft, Meta, Amazon) is pouring money into AI: are those capex dollars actually turning into operational wins and new revenue today, or mostly bets on a future payoff?

Kate Leaman: There’s a clear mix of both operational wins and future bets. On the one hand, the large players are already seeing revenue uplift linked to their AI infrastructure and offerings.

For example, their cloud and advertising businesses are leveraging AI capabilities to drive growth.

But on the other hand, their capital expenditures (capex) are growing much faster than revenue, so a large portion of the spend is still making its way into future payoff territory.

Aggregated capex by major ‘hyperscaler’ tech firms is approaching record levels relative to their operating cash flow, for example, one estimate puts aggregate capex at roughly 60% of operating cash flow for Amazon, Google/Alphabet, Microsoft, and Meta.

Analysts flag that unless revenue increases markedly from the AI‑investments, many of these companies will be reinvesting nearly all of their free cash flow into infrastructure in just a few years.

On the win side, we see incremental revenue from AI‑driven cloud services, generative AI features in products, and improved operating efficiency; for instance, some smaller cloud/AI vendors show that AI capex is paying off already.

Invezz: Alphabet says it has a record $155 billion cloud backlog driven by AI. Does that point to durable enterprise adoption or a risk of overstating near‑term growth?

Kate Leaman: This is a nuanced one. The $155 billion backlog for Alphabet Inc.’s cloud business (via Google Cloud) is an encouraging signal of strong enterprise demand, but it also carries caveats.

Supporting the durable adoption thesis is that the reported backlog grew significantly, and analysts at UBS described it as a ‘strong growth’ signal.

Alphabet’s commentary emphasises that this backlog is tied to its ‘full‑stack’ AI approach (infrastructure, AI models, enterprise solutions), which gives it differentiation and suggests more than just hype.

So the backlog shows promise of durable enterprise adoption.

But on the risk side, a backlog is not the same as revenue recognised today; the path from backlog to realised revenue and margin is not always smooth or guaranteed.

Invezz: Apple’s often called a late entrant to AI: what concrete moves should it make to close the gap, and how badly could a slow AI strategy hurt its competitive edge?

Kate Leaman: For Apple, the situation is two‑fold: its strategy is distinct (privacy‑first, on‑device intelligence), which gives it some advantages, but the slower pace also raises risks.

Concrete moves Apple should make include:

• Better integration of AI across its ecosystem beyond just privacy and on‑device features.
• Accelerating development of its AI assistants (e.g., improvements to Siri) and making them more competitive with generative AI offerings from rivals.
• Pursuing strategic acquisitions or partnerships to bring in generative AI and multimodal capabilities faster.

If Apple falls behind on AI momentum, it risks losing mind‑share among early adopters and developers who are being drawn to competitors with more visible AI leadership.

This is also the worry that its service business growth will slow if rival platforms (cloud‑centric, AI‑centric) capture the ecosystem of apps and tool, as well as concerns of reduced margin growth over time if hardware remains strong but software/services growth lags.

Invezz: Meta’s AI‑powered ad platform is fuelling growth again. How resilient is that model if regulators tighten rules around user data and privacy?

Kate Leaman: For Meta, the resurgence via its AI‑driven advertising platform demonstrates strong potential: its use of generative AI and improved ad personalisation are giving it a boost.

But the model is exposed to significant regulatory and privacy risk.

Resilience factors include Meta embedding AI in recommendation engines and ad targeting; for example, it announced that user interactions with its AI assistant will help personalise content and ads.

It also has a vast user base and data assets, which give it scale advantages.

Risk factors include that new privacy regulations globally (e.g., in the EU, some U.S. states) are increasing scrutiny over how platforms use personal data and how transparent they are.

Invezz: The Magnificent Seven now account for 37 % of the S&P 500. How should investors chase AI upside without increasing concentration risk in their portfolios?

Kate Leaman: The fact that the ‘S&P 500 large‑cap tech group’ (often labelled the ‘Magnificent Seven’) now holds such a large share of the market means that chasing AI upside via only the major tech names carries concentration risk.

Here are balanced ways investors can manage this:

• Using tactical portfolio risk management. This could include rebalancing regularly or using options/hedges, if an investor believes valuations are extended.”
• Diversification strategies include pairing exposure to large AI‑led tech companies with selected beneficiaries outside mega‑caps e.g., industrial firms applying AI, healthcare companies deploying AI, and enterprise software vendors. This spreads the risk of a setback in any one large company.
• Using thematic funds or ETFs focused on AI but with broader sectoral or company‑size exposure (not just mega‑caps)
• Including value or dividend‑growth stocks that benefit from AI indirectly. For example, companies whose operations become more efficient via AI, but are less exposed to high‑multiple valuations

Invezz: As AI shifts from trend to necessity, what are the handful of factors that will separate companies that sustain growth and margins from those that don’t?

Kate Leaman: The main differentiators include:

• Talent, ecosystem, and partnerships: success will favor firms that build or attract top AI talent, create developer ecosystems, integrate partners, and iterate rapidly.
• Capital‑allocation discipline: firms that invest sensibly in AI infrastructure with an eye on return on capital (not just chasing the ‘AI buzz’) will have a stronger stance.
• Enterprise adoption and diversification of revenue: it’s not enough to build models; companies must win enterprise customers, embed AI into business processes, and diversify across consumer, enterprise, and platform layers.
• Privacy‑aware and scalable AI models: as regulation and public scrutiny intensify, companies that build AI with trust, compliance, scalability, and efficiency in mind will outperform those treating AI only as marketing.
• Regulatory navigation: new and evolving regulations (data privacy, AI governance, antitrust implications) mean that companies proactively shaping and adapting to compliance and governance, rather than reacting, will have a competitive advantage.
• Cost/margin management: infrastructure, compute, and hardware costs are rising rapidly. Companies that can maintain margins while scaling AI will succeed – those whose AI drives revenue but burns cash will falter.
• Product differentiation and integration: AI is now table‑stakes; the winners will embed it deeply into differentiated products or services, rather than bolt‑on superficial features.

The post Interview: 'Big Tech's AI capex is growing much faster than revenue,' warns AvaTrade’s Kate Leaman appeared first on Invezz