Artificial intelligence has become a major investment priority for the world’s largest technology companies. But building the infrastructure needed to power AI is expensive, and investors are paying closer attention to how that spending affects profitability and cash flow.
A market sell-off on October 22, 2026, highlighted those concerns. Technology stocks came under pressure as investors assessed the cost of AI expansion and the financial commitments required to sustain it.
Reuters reported that Meta shares fell 7% during the session, while Nvidia declined 4%. The Nasdaq Composite dropped 2.6%.
The reaction illustrates a growing question for the technology industry: how quickly can companies turn their investments in AI infrastructure into sustainable financial returns?
Why is AI infrastructure so expensive?
Running advanced AI systems requires more than developing a model. Companies need computing hardware, data centres, networking equipment, electricity and the infrastructure to deliver services to customers.
These investments can require substantial upfront spending, while the revenue they eventually generate may take longer to materialise.
Technology companies are therefore making decisions about how quickly to expand capacity, how to finance new infrastructure and how much spending their existing businesses can support.
The challenge is particularly important for businesses investing ahead of demand. If customer adoption or revenue growth falls short of expectations, the cost of maintaining that infrastructure can put pressure on financial performance.
What do Meta’s financial figures tell us?
Meta’s third-quarter 2026 financial disclosures provide a useful example of the scale involved.
The company reported third-quarter revenue of $64.60 billion. It also recorded negative free cash flow of $2.12 billion, compared with positive free cash flow of $10.63 billion in the same quarter a year earlier.
Meta’s capital expenditure outlook for 2026 was between $135 billion and $145 billion, with the lower end of the range raised in its updated forecast.
These figures show the financial demands associated with the company’s spending plans. They do not, by themselves, establish whether its AI investments will ultimately succeed or fail.
Investors must also consider revenue growth, operating performance, the expected useful life of infrastructure and the potential returns from AI products.
Sources: Meta’s official third-quarter financial results and Reuters’ reporting on AI infrastructure spending.
Why are investors paying closer attention?
Large investments are easier to justify when businesses can demonstrate a credible path to higher revenue, improved productivity or lower operating costs.
For AI, that calculation is complicated by the pace of technological change. Companies may need to invest in new hardware and infrastructure even as the models and applications they support continue to evolve.
Investors are consequently looking beyond announcements about new AI products. They want to understand whether demand is strong enough to justify the associated spending and whether companies can maintain healthy cash generation while expanding.
This does not mean the market has abandoned AI. Rather, expectations about future growth are being weighed against the costs required to achieve it.
What does this mean for smaller businesses?
Most small businesses do not need to build data centres or purchase specialised computing infrastructure to benefit from AI. They can access many capabilities through existing software subscriptions and cloud services.
Nevertheless, the same financial principle applies: technology should produce value that justifies its cost.
A small business considering an AI customer-support tool, for example, should compare the subscription and implementation costs with the time saved, service quality and measurable business outcomes.
The most expensive tool is not necessarily the most useful, and the newest technology is not automatically the right choice.
Businesses should also account for less obvious expenses, including employee training, integration, human review and the time required to maintain reliable workflows.
What to watch next
Several indicators will help determine whether major AI investments are delivering the expected returns:
- Revenue from AI products: Are customers paying for AI capabilities at a meaningful scale?
- Capital expenditure: Are infrastructure budgets continuing to rise, or are companies gaining more control over spending?
- Free cash flow: Can businesses finance expansion while maintaining financial flexibility?
- Productivity gains: Are AI tools generating measurable improvements for customers and employees?
- Competition: Are lower-cost models and alternative providers changing the economics of AI deployment?
These indicators matter more than any single day’s share-price movement.
The bottom line
AI infrastructure is becoming a major financial commitment for the technology industry. Meta’s spending plans and the October 22 market reaction show why investors are scrutinising the relationship between investment, cash flow and future returns.
The long-term opportunity in AI remains significant, but the economics will differ across companies and applications. For investors and business owners alike, the key question is not simply how much is being spent on AI. It is whether that spending creates enough lasting value to justify the cost.