OpenAI is ramping up its financial outlook, now expecting to burn a staggering $115 billion by 2029—a massive increase of $80 billion over previous forecasts. This surge in spending underscores the company’s aggressive push to scale its AI infrastructure and achieve leadership in the AI space.Reuters
Background: Why Costs Have Skyrocketed
The dramatic jump in projected spending comes on the heels of soaring infrastructure demands. Within 2025 alone, OpenAI anticipates spending more than $8 billion, up by $1.5 billion from earlier estimates. This escalation reflects its growing reliance on compute-intensive operations to support advanced AI, especially its popular ChatGPT services.
OpenAI’s updated financial model includes planned capital expenditure that more than doubles year-on-year, with cash burn expected to reach $17 billion in 2026, $35 billion in 2027, and $45 billion in 2028.
To manage this escalating cost structure, the company is investing strategically in internal capabilities—developing its own AI chips in partnership with Broadcom—and expanding its physical infrastructure through major collaborations, notably the 4.5-gigawatt data center build with Oracle as part of the larger $500 billion “Stargate” initiative.
Strategic Moves: Chips and Data Centers
OpenAI’s chip development strategy—co-designing its first AI processor with Broadcom—is a pivotal step toward controlling compute costs and securing supply chains critical to its expansion. These chips are intended for internal use only.
In addition to chip fabrication, OpenAI continues to deepen ties with cloud providers. Its infrastructure build with Oracle under the Stargate initiative is supplemented by computing contracts with Google Cloud. This phase of intensive infrastructure investment is designed to accommodate ever-growing AI demands and reduce reliance on external vendors like Nvidia.
Industry Context: Why the Spending Matters
OpenAI’s expenditure forecast signals a bold bet on AI’s future. Analysts estimate that global data center capital expenditure for AI-related infrastructure could hit $1.2 trillion by 2029, growing at a 21% compound annual growth rate—highlighting how OpenAI’s efforts reflect a broader infrastructure boom.
Such investments not only prepare OpenAI to scale its AI models, but also position it to offer infrastructure capabilities to external customers—an emerging revenue stream highlighted by CFO Sarah Friar’s comments about building trillion-dollar-scale data centers.
What’s Next: Execution and Impact
- Infrastructure rollout: Continued expansion under the Stargate initiative remains key—especially in building data centers in India and beyond to support global AI deployment
- Profit timeline: OpenAI still does not expect to turn cash-flow positive until 2029, balancing massive spending with exponential revenue growth
- Financial strategy: With capital burn ramping up sharply, OpenAI’s funding strategy and cost control methods (chips, infrastructure ownership) will play a crucial role in navigating this scale-up phase.
