The rapid expansion of artificial intelligence is creating a new challenge for the technology industry: finding enough money to finance its growth. Intel CEO Lip-Bu Tan and Broadcom CEO Hock Tan have warned that access to capital and the rising cost of financing could become major obstacles to the AI boom.
Speaking at the DBS Asia Future Forward Conference in Singapore, the executives highlighted concerns about the enormous investments required to build AI infrastructure. While demand for computing power remains strong, companies must ensure they can afford the data centres, advanced chips and electricity infrastructure needed to support that demand.
Their comments suggest that financing could become a bigger concern than some of the hardware and energy shortages currently affecting the industry.
Why Capital Is Becoming a Major Concern for AI Companies
The growth of generative AI has pushed technology companies to invest heavily in computing infrastructure. Training and operating advanced AI models requires powerful processors, large amounts of memory, data centres and reliable electricity supplies.
These investments can cost billions of dollars and may take years to deliver financial returns.
Broadcom CEO Hock Tan said access to capital could become the industry’s biggest constraint as companies continue expanding their AI capabilities. He noted that the sector has already faced several supply bottlenecks, but financing could prove more difficult to overcome.
Intel CEO Lip-Bu Tan expressed similar concerns, pointing to the cost of capital and the need for companies to demonstrate a clear return on investment.
When borrowing becomes more expensive, businesses may find it harder to finance large infrastructure projects. Investors may also become more cautious about supporting projects whose profitability depends on future AI demand.
Rising Interest Costs Could Affect AI Infrastructure Spending
Building AI infrastructure involves substantial upfront spending. Companies need to secure land, construct data centres, purchase specialised chips and arrange sufficient power capacity before many projects can generate revenue.
Higher borrowing costs can increase the total expense of these projects and make financial planning more challenging.
For example, a company planning a large AI data centre must consider whether future revenue from cloud computing, AI services or enterprise customers will justify the initial investment.
If financing costs rise or demand grows more slowly than expected, the project could become less profitable than originally planned.
This does not necessarily mean AI investment will stop. However, companies may become more selective about which projects they fund, how quickly they expand and how they allocate capital between different technologies.
Semiconductor and Power Shortages Could Gradually Ease
Despite their concerns about financing, the executives suggested that some of the industry’s existing infrastructure constraints could improve over time.
Broadcom’s Hock Tan said the technology sector has repeatedly encountered supply bottlenecks over the past two years, but some have eased as manufacturers expanded capacity and businesses adapted.
He suggested that today’s memory shortages might look different a year later. He also indicated that wafer supply constraints could ease around 2028 or 2029, while electricity availability could improve as additional infrastructure comes online.
These developments could help technology companies build and operate AI systems more efficiently.
However, improvements in chip production and power availability would not automatically resolve financing challenges. Even when equipment becomes easier to obtain, companies still need sufficient funding to purchase it and operate the facilities that use it.
AI Computing Demand Could Remain Strong Through 2030
Both executives remain optimistic about near-term demand for AI computing infrastructure.
Hock Tan said demand for computing capacity currently exceeds what the industry can supply. He expects this imbalance to continue through 2029 and potentially into 2030.
Lip-Bu Tan also indicated that the industry is unlikely to experience a significant slowdown over the next two to three years, supported by the growing adoption of AI applications and demand for computing power.
This outlook reflects the expansion of AI across cloud services, enterprise software, software development and other business operations.
As companies introduce AI-powered products and automate more tasks, they may require additional computing resources to develop, deploy and maintain these systems.
Nevertheless, strong demand forecasts do not guarantee that every infrastructure investment will produce attractive returns. The financial performance of individual projects will depend on costs, utilisation rates and the revenue generated by the services they support.
Will AI Investments Generate Enough Returns?
A central question for the industry is whether the commercial benefits of AI will justify the huge amounts of money being invested.
Companies are spending heavily on infrastructure in anticipation of continued growth in AI usage. However, investors increasingly need evidence that these investments can translate into higher revenue, lower operating costs or entirely new business opportunities.
Broadcom’s Hock Tan warned that demand could weaken if advanced AI models fail to improve as expected or if progress in their capabilities begins to slow.
Intel’s Lip-Bu Tan also stressed the importance of demonstrating measurable returns from AI investments.
For businesses adopting AI, this means moving beyond experimentation and identifying applications that deliver practical benefits. These could include improving employee productivity, reducing processing costs, accelerating software development or creating new products and services.
If companies cannot demonstrate these benefits, they may become more cautious about increasing their AI budgets.
What the Capital Challenge Means for the AI Industry
The executives’ comments highlight a potential shift in the AI infrastructure market. In the early stages of the boom, attention focused heavily on the availability of advanced processors, memory chips and electricity.
As manufacturers and infrastructure providers work to address those constraints, financing and profitability may become more important factors in determining how quickly the industry can expand.
Companies with strong balance sheets, reliable funding and clear business models could be better positioned to continue investing. Others may need to delay projects, seek new financing arrangements or focus on infrastructure that can generate revenue more quickly.
The challenge could also influence competition among AI developers and cloud providers, since large-scale infrastructure investments require both technical expertise and substantial financial resources.
What Comes Next?
AI computing demand remains strong, but sustaining the industry’s growth will require more than expanding chip production and building additional data centres.
Technology companies will need access to affordable capital and convincing evidence that their investments can produce long-term economic value. Progress in AI applications, customer adoption and infrastructure efficiency will all influence future spending decisions.
Conclusion
Intel and Broadcom’s CEOs believe financing could become one of the biggest challenges facing the AI industry. While demand for computing power is expected to remain strong in the coming years, rising capital costs and uncertainty about returns could influence how quickly companies expand their infrastructure. The next phase of the AI boom may depend not only on technological progress, but also on whether businesses can turn their investments into sustainable profits.