NextFin News - In a move that has sent ripples through Silicon Valley and global capital markets, OpenAI has officially halved its long-term infrastructure investment projections. According to The Chosun Daily, the San Francisco-based AI pioneer informed investors on February 21, 2026, that it now plans to allocate approximately $600 billion toward AI infrastructure development by 2030. This figure represents a staggering retreat from the $1.4 trillion ambitious spending roadmap previously championed by the company’s leadership. The decision comes as U.S. President Trump’s administration continues to emphasize domestic industrial efficiency and as institutional investors grow increasingly wary of a potential "AI bubble" characterized by astronomical capital expenditures without immediate commensurate returns.
The timing of this recalibration is critical. OpenAI is currently navigating the final stages of a pre-IPO funding round exceeding $100 billion, which includes a significant $30 billion direct investment from Nvidia. While the company’s valuation is expected to climb to $830 billion upon the completion of this round, the reduction in planned spending suggests a strategic pivot toward sustainability. According to CNBC, OpenAI’s Chief Financial Officer Sarah Pryor recently confirmed that the company’s annual revenue surpassed $20 billion last year—a threefold increase from the $6 billion recorded in the previous period. Despite the spending cuts, internal estimates now project 2030 annual sales to reach $280 billion, surpassing the earlier target of $200 billion.
This divergence—lowering investment while raising revenue guidance—reveals a fundamental shift in the generative AI business model. The initial phase of the AI race was defined by "scaling laws," the belief that more data and more compute would inevitably lead to more capable models. However, the industry is now entering a "diminishing returns" phase where the marginal cost of adding another 100,000 GPUs may no longer be justified by the incremental performance gains. By slashing its infrastructure budget by $800 billion, OpenAI is signaling that the path to artificial general intelligence (AGI) may require algorithmic efficiency and architectural innovation rather than just brute-force hardware expansion.
From a macroeconomic perspective, the move reflects a broader cooling of the "compute-at-all-costs" era. The high-interest-rate environment and the 15% global tariff hikes recently implemented by U.S. President Trump have increased the cost of importing specialized components and maintaining global data center footprints. Analysts suggest that OpenAI’s decision is a preemptive strike to protect its margins ahead of its highly anticipated initial public offering. By demonstrating fiscal discipline, the company aims to distance itself from the "dot-com" comparisons that have plagued the sector throughout 2025 and early 2026.
The shift toward business-to-business (B2B) services and ad integration within ChatGPT further underscores this drive for profitability. Pryor’s reporting of $20 billion in revenue indicates that enterprise adoption of customized AI models is scaling faster than consumer subscriptions. This B2B pivot allows OpenAI to offload some infrastructure burdens onto corporate clients who run models on their own private clouds, thereby reducing the capital intensity of OpenAI’s own balance sheet. Furthermore, the $30 billion investment from Nvidia suggests a deepening vertical integration, where hardware providers are becoming equity partners to secure long-term demand, even as the total volume of that demand is revised downward.
Looking ahead, OpenAI’s retreat from the trillion-dollar investment threshold likely marks the end of the "hyper-expansion" phase of the AI cycle. As the company prepares for its IPO, the focus will shift from how many chips it can acquire to how much revenue it can extract per watt of power consumed. If OpenAI can achieve its $280 billion revenue target with only $600 billion in infrastructure spend, it will prove that AI is a viable, high-margin industry rather than a capital-intensive black hole. However, if competitors like Anthropic or xAI continue to spend aggressively, OpenAI may find itself forced to choose between fiscal conservatism and technological dominance in the race toward 2030.
Explore more exclusive insights at nextfin.ai.
