Big Tech to Spend 320 billion USD on AI in 2025

Inspired by: Morris, S. & Uddin, R. (2025). Big Tech lines up over $300bn in AI spending for 2025. Financial Times. https://www.ft.com/content/634b7ec5-10c3-44d3-ae49-2a5b9ad566fa

Block 1

Word Count - 769

Big Tech to Spend 320 billion USD on AI in 2025  

Spending from the "Magnificent Seven" dwarves that of the rest of the stock market, and is increasing with Amazon, Meta, Microsoft, and Alphabet forecasting to spend upwards of a combined $320 billion on artificial intelligence (AI) in 2025. From revolutionizing cloud computing to enhancing productivity tools and transforming advertising, AI has potential to unlock new revenue streams and drive efficiency. While large investments into AI are framed as essential for securing long-term competitive advantages, they have sparked questioning from investors. Are these expansive capital expenditures a path to shareholder value creation, or do they risk destroying value in pursuit of an uncertain future?   

Pulled from: Morris, S. & Uddin, R. (2025). Big Tech lines up over $300bn in AI spending for 2025. Financial Times. https://www.ft.com/content/634b7ec5-10c3-44d3-ae49-2a5b9ad566fa

How Will Increased Investment into AI Influence Shareholder Value?  

There are two trails of thought for these investments: Either this aggressive approach will create important infrastructure and technological breakthroughs, creating shareholder value, or such expensive investment will destroy shareholder value as cheaper AI competitors like Deepseek catch up. I'm in this second camp. Such heavy investment is a waste of capital that could be better spent increasing capabilities elsewhere until the technology develops further. 

I have firsthand seen the rapid increase of AI capabilities and have been a frequent user of the newest advancements, taking advantage of text-based AI such as ChatGPT and DeepSeek as well as image creating platforms such as ComfyUI, and StableDiffusion. AI has been observably growing at an exponential rate and diminishing returns are nowhere in sight. The Magnificent 7 are fighting for market share in the AI sphere. By investing in creating AI advancements and infrastructure, these companies can gain first mover advantages, gaining access to technologies that other organisations don't yet have. Additionally, investment will allow these companies to attract and maintain top talent. With the AI job sector being so new, talented employees are likely more scarce than these companies would like, so it's important to attract and keep these scarce human resources.   

But, I believe that grasping for control of this fast-growing market is futile and the organisations who prioritize strategic use of AI rather than development of it will create more shareholder value. As seen with the release of DeepSeek and subsequent destruction of $600 billion of Nvidia’s market value, first mover advantages are hard to sustain within the AI environment due to increasing availability of cheap alternatives. It may be a more efficient use of time and money to wait for these technologies to grow externally before investing. 

DeepSeek's rise underscores an apparent truth: agility trumps scale in fast-moving tech sectors. Shareholder value is increasingly being created in companies that can pivot swiftly and prioritize optimization over maximisation.  

Relating to Theory

The “Value Action Pentagon” (Rapp et al., 2011) offers a framework to evaluate how these investments will affect shareholder value. To best create shareholder value, investment into AI must satisfy the 5 parts of the value action pentagon. The following is an analysis of heavy investment into Ai using this framework.  

The value action pentagon 

1. Increase returns on existing capital: Tech giants must ensure AI-driven tools boost productivity or pricing power.  

2. Invest in positive-spread projects: AI initiatives must generate returns above the cost of capital. 

3. Divest from underperforming assets: Redirecting resources from legacy projects to AI could refine capital allocation, mirroring Netflix’s 2011 pivot from DVDs to streaming. We don't know where the cash for these investments is coming from. Should it come from underperforming ventures, investing in AI may be a better use of that money. 

4. Extend the planning horizon: AI’s payoff may take decades. However, overreliance on distant returns risks hurting shareholder value in the short term. 

5. Lower the required return: Partnerships (e.g. Microsoft partnering with OpenAI) could reduce risk. Yet, as DeepSeek's rise shows, agility often trumps scale - highlighting the need for operational flexibility. 

 

The Bottom Line  

AI is transformative, but transformation ≠ profitability. As the introduction of DeepSeek showed, markets now prefer agility - the ability to adapt, optimize, and pivot - over sheer scale. And shareholders are aware agility is possible, therefore expecting it from the companies they invest in.  

I believe it is important to spend smart, not big. Investing in AI is not a bad idea with myself and many shareholders seeing AI as the future. But, it's important to understand that companies that don't balance ambition with discipline may find their financial statements covered in wasted expenditure, wondering how the next DeepSeek bested them with such fewer resources. 

 

References 

Rapp, M. S., Schellong, D., Schmidt, M., & Wolff, M. (2011). Considering the shareholder perspective: Value-based management systems and stock market performance. Review of Managerial Science, 5, 171-194.