The AI boom is entering a new phase. The biggest question is no longer just how many people use AI, but how much businesses are willing to pay for it.
OpenAI is reportedly targeting $70 billion or more in annualised revenue by the end of 2026, driven largely by growth in its enterprise business. Reports put its annualised revenue at approximately $50 billion at the end of September, making the year-end target an ambitious goal rather than an achieved result.
That is a significant development for the AI industry. It also raises an interesting question for stock investors: if AI companies are generating revenue at this scale, which publicly traded businesses stand to benefit most?
My attention is on three names: Microsoft, Nvidia and Oracle. Each plays a different role in the AI ecosystem, and each faces a different set of risks.
🔥 1. The real story is enterprise AI
For a while, much of the AI conversation focused on ChatGPT’s popularity, new models and the race to develop increasingly capable technology.
But enterprise adoption could be an even more important long-term story.
Businesses are exploring how AI can help employees write and analyse documents, automate repetitive tasks, develop software, support customer service and process large amounts of information. Governments and large organisations are also investigating AI applications that require secure infrastructure and greater control over their data.
The commercial opportunity is potentially enormous because companies may be willing to pay recurring fees for tools that save time, improve productivity or help them operate more efficiently.
However, there is a difference between experimenting with AI and deploying it across an entire organisation.
Businesses will continue spending only if these tools provide measurable value. If the productivity gains are difficult to demonstrate, some customers may reduce their spending or delay wider adoption.
The next stage of the AI race will depend on turning impressive technology into products that customers repeatedly pay for.
💻 2. Microsoft: The enterprise distribution advantage
Microsoft is an interesting company to watch because it already has deep relationships with businesses around the world.
Its products are embedded in everyday corporate operations, from workplace software to cloud computing. This gives Microsoft several ways to introduce AI services to existing customers.
Its relationship with OpenAI also connects it to the development and commercialisation of AI technology.
The investment case is not simply that Microsoft has exposure to OpenAI. It is that Microsoft can incorporate AI into its own products, distribute those services through established channels and potentially generate additional revenue from customers who already use its ecosystem.
For investors, the key questions are whether customers will pay for premium AI features, whether usage will expand and whether the additional revenue will justify the infrastructure costs.
Microsoft also has to invest heavily in data centres, computing capacity and related infrastructure. Strong demand is encouraging, but the cost of meeting that demand matters.
What I would watch: Azure growth, commercial AI adoption, cloud margins and evidence that AI services are contributing meaningfully to Microsoft’s financial performance.
🟢 3. Nvidia: The computing engine behind the boom
Nvidia occupies a different position.
Rather than selling AI applications directly to most end users, Nvidia supplies the computing technology that enables many AI systems to operate.
Training and running sophisticated AI models require substantial computing resources. As AI adoption expands, demand for processors, networking equipment and complete computing systems can increase.
This creates a potential benefit for Nvidia if OpenAI and other AI developers continue expanding their computing capacity.
However, investors should not assume that every dollar of AI revenue translates directly into Nvidia revenue. AI companies spend money across several areas, including cloud services, staff, research, software development and infrastructure.
Furthermore, Nvidia faces competition, the development of custom chips by large technology companies and the possibility that AI computing becomes more efficient over time.
Efficiency is an important factor. If businesses can accomplish more with fewer computing resources, the amount of infrastructure required for each AI task could decline. At the same time, lower costs could encourage much wider AI usage, potentially increasing total demand.
The balance between these forces will help determine the industry’s long-term growth.
What I would watch: Data-centre demand, gross margins, customer spending plans and whether computing demand remains strong as AI models become more efficient.
☁️ 4. Oracle: The infrastructure opportunity
Oracle is another company worth following because AI developers need substantial cloud infrastructure to build, train and operate their systems.
Oracle provides cloud computing and database services, and its infrastructure business gives it exposure to demand for AI computing capacity.
If AI companies continue expanding, they may need more computing resources, data-centre capacity and long-term infrastructure commitments.
That creates opportunities for infrastructure providers, but it also introduces financial risks.
Building data centres requires substantial capital. Companies must spend on equipment, facilities, networking and power before all the expected revenue necessarily arrives.
Long-term contracts can provide greater visibility over future demand, but investors should still examine the cost of fulfilling those commitments.
A large backlog may look impressive, yet the timing of revenue recognition, capital expenditure and cash generation remains important.
What I would watch: Cloud growth, infrastructure demand, capital expenditure, debt and free cash flow.
The key question is whether Oracle can turn its growing role in AI infrastructure into sustainable financial returns.
📊 5. The $70 billion figure needs context
There is one important detail investors should not overlook.
Annualised revenue is not the same as revenue actually earned over a full financial year.
It is generally calculated by taking revenue generated over a shorter period and projecting it across 12 months. This can be useful for fast-growing businesses, but it is not a guarantee of future sales.
OpenAI’s reported figures have also attracted attention because different companies may calculate or present revenue run rates differently.
Recent reporting has described approximately $50 billion in annualised revenue at the end of September, alongside a target of reaching or exceeding $70 billion by year-end.
Investors should therefore distinguish between the current reported run rate, the year-end target and the revenue ultimately recognised in financial statements.
This matters because a headline number can influence market expectations even when it does not tell the whole financial story.
Revenue alone also does not reveal profitability.
AI companies must pay for computing infrastructure, electricity, employees, research and product development. Growing sales are encouraging, but the cost of generating those sales determines whether the business model can become financially sustainable.
For investors, I would want to understand both sides of the equation: how quickly revenue is growing and how much it costs to support that growth.
⚠️ 6. What could derail the AI investment story?
There are several risks worth considering.
First, spending may grow faster than profits. AI infrastructure is expensive, and strong demand does not automatically guarantee attractive returns on invested capital.
Second, competition is increasing. AI developers face pressure from established technology companies, emerging competitors and open-source alternatives. Competition could reduce prices and put pressure on margins.
Third, expectations may already be high. A company can deliver strong results and still see its shares fall if investors expected even better performance.
Fourth, infrastructure demand may fluctuate. Technology companies are making enormous investments based on expectations of future AI usage. If adoption grows more slowly than expected, some projects could be delayed or spending plans reduced.
Finally, AI efficiency could change the economics of the industry. Better models may reduce the computing resources needed for certain tasks, even as cheaper AI services attract more users.
These risks do not mean the AI opportunity is disappearing. They mean investors need to distinguish between a growing industry and a stock that offers an attractive return at its current valuation.
🎯 My investment takeaway
I find the enterprise AI story particularly interesting because it moves the discussion beyond consumer excitement and towards business spending.
OpenAI’s revenue ambitions suggest that the market for AI services could be substantial. Microsoft has the advantage of enterprise distribution, Nvidia supplies critical computing technology, and Oracle is positioned to benefit from demand for cloud infrastructure.
But these are different businesses, with different valuations, financial commitments and risk profiles.
I would not buy any of them simply because OpenAI reports strong revenue growth. Instead, I would look for evidence that AI demand is translating into sustainable revenue, healthy margins and cash flow.
The biggest winner in AI may not necessarily be the company with the most impressive revenue headline. It could be the company that turns AI demand into durable profits while managing its investment costs.
Which would you rather invest in for the next three years: Microsoft for enterprise AI, Nvidia for computing power, or Oracle for AI infrastructure?
Disclaimer: Not financial advice. Do your own research before investing.
Comments