As competition in AI intensifies, achieving the next breakthrough is becoming increasingly expensive. Amid this race, established tech companies are leading research and development (R&D) spending, fueling innovation across the AI ecosystem and fostering competition. 

AI Development Requires Significant Resources

As we’ve previously written, advancing to the next level of AI requires unprecedented investment across the AI stack, a barrier that both fuels competition and increases innovation.

— According to the United Nations’ (UN) Technology and Innovation Report 2025, the cost of training frontier AI models has increased 2.4 times per year since 2016, with more than half of the development cost attributed to hardware. 

— Similarly, an article from The Wall Street Journal noted that “outsize spending” is “necessary for machine-learning systems to reach artificial general intelligence, or AGI, a state in which they are smarter than humans.”

As AI systems scale, success is increasingly determined by access to capital-intensive infrastructure. This shift is turning AI development into a resource-constrained competition, where physical scale and infrastructure matter as much as technical capability. Furthermore, the footprint required for successful AI development is expanding rapidly. 

— Established tech companies poured tens of billions of dollars into AI infrastructure and data centers last year, investments they describe as “necessary for the internet’s next era,” according to CNN.

— The scale of this buildout is set to accelerate. Deloitte’s 2025 AI Infrastructure Survey estimates that power demand from AI data centers in the U.S. will grow more than thirtyfold, reaching 123 gigawatts by 2035. The report underscores that the evolving scope and scale of AI infrastructure development continue to increase “the investment stakes of building out capacity for data centers, power generation, and manufacturing to trillion-dollar levels.” Cloud infrastructure providers play a central role in this buildout, offering scalable compute resources, including custom AI chips, that allow companies of all sizes to access cutting-edge AI capabilities without building proprietary data centers.

Large-Scale Investments Lower Barriers for Smaller AI Innovators

Substantial investments by well-resourced companies can increase competition by lowering barriers for smaller innovators. A healthy AI ecosystem depends on both large-scale infrastructure providers and agile startups.

— CNBC’s interview with Alice Bentinck, founder of the startup Entrepreneurs First, highlighted that “large AI companies are complementary to AI start-ups” and noted that the benefits startups gain from larger firms ultimately strengthen the entire industry.

— Bentinck stated that “[bigger platforms] are producing absolutely incredible models, and there’s an exciting layer [where] AI startups are being built on top of those. As those models improve, it also improves the kind of products and services those [startups] can deliver.” 

— More examples come as large companies invest directly in small AI innovators that require significant capital. In 2025, established tech companies played a significant role in backing AI startups, including Amazon‘s nearly $8bn cumulative investment in Anthropic—one of the largest AI startup investments ever made—alongside its investment in multiple other AI companies.

Ultimately, the massive R&D spending and investment activity of established technology firms advances core AI capabilities and works to enable all companies to participate in a competitive, innovative, and dynamic AI ecosystem.