OpenAI reportedly spent $34 billion in 2025 as it expanded computing capacity, developed new models and prepared for a possible IPO. Figures reported by the Financial Times and covered by Reuters suggest around $19 billion went towards research and development, while nearly $6 billion was spent on sales and marketing.

For investors, the main question is where that money is going. The spending could support chips, cloud computing and data-centre infrastructure, while increasing pressure on smaller AI companies and software businesses.

Winners

AI chips and custom silicon

Advanced AI models require enormous processing power. Nvidia remains the leading supplier of AI accelerators, while AMD could benefit as customers seek alternatives.

Broadcom may gain from demand for custom AI chips and networking technology. Larger computing clusters require more processors.

Names: $NVDA (Nvidia), $AMD (Advanced Micro Devices), $AVGO (Broadcom)

Cloud and AI infrastructure

OpenAI needs more computing capacity than it can build alone, creating opportunities for cloud providers and specialist GPU infrastructure companies.

Oracle could benefit from major AI contracts and the Stargate build-out. Amazon may gain through Amazon Web Services. CoreWeave offers a concentrated way to trade rising GPU demand.

Names: $ORCL (Oracle), $AMZN (Amazon), $CRWV (CoreWeave)

Data-centre networking, cooling and power

Arista supplies high-speed networking, Vertiv provides cooling and power management, and Eaton supplies electrical distribution equipment.

AI servers use more electricity and produce more heat than conventional servers. As cloud companies expand capacity, demand for these systems may rise.

Names: $ANET (Arista Networks), $VRT (Vertiv), $ETN (Eaton)

Losers

Smaller AI software companies

OpenAI’s $19 billion research budget shows the challenge facing smaller AI companies. OpenAI can improve models, lower prices, enter new markets and spend heavily to win customers.

Smaller companies may still succeed in specialist areas, but investors will demand stronger recurring revenue, improving margins and defensible technology.

Names: $AI (C3.ai), $SOUN (SoundHound AI), $BBAI (BigBear.ai)

Established software providers

OpenAI’s expanding capabilities could pressure software companies selling specialised tools at premium prices.

Salesforce must show that AI agents create new revenue. Adobe faces competition from generative image and video tools, while Intuit could see more bookkeeping and tax work automated.

These companies have valuable data and customer relationships, but AI may make some features less distinctive and pressure pricing.

Names: $CRM (Salesforce), $ADBE (Adobe), $INTU (Intuit)

Technology consulting and outsourced digital work

Coding agents and automated workflows could reduce the billable hours needed for development, testing and support.

Accenture, EPAM and Globant may benefit from helping clients adopt AI, but they must move towards higher-value consulting. Businesses relying on large technical teams could face pressure on utilisation, pricing and hiring.

Names: $ACN (Accenture), $EPAM (EPAM Systems), $GLOB (Globant)

#StockMarket #Trading #Investing #DayTrading #SwingTrading #OpenAI #AIStocks #TechStocks #Nvidia #Oracle #Amazon #DataCenters #CloudComputing #Semiconductors #SoftwareStocks #IPO #WallStreet