With big tech shelling out big bucks for AI dominance, here’s a look at how much they are spending, what they are spending it on, and whether it could help strengthen their market position in AI.

Microsoft

In its fourth quarter FY2024 earnings call, Microsoft said that out of the $19 billion it has reported as capital expenditures (capex), most of the spending was related to AI. While half of its capex was spent on infrastructure such as building data centres, the remaining amount was used to purchase CPUs and GPUs to enhance cloud computing as well as AI processing power.

Besides shoring up its own AI offerings, Microsoft is also one of the biggest investors in OpenAI – the AI startup that created ChatGPT. Though OpenAI has emerged as a major player in the AI race, the company’s profitability has been brought into doubt as its operating costs could reach $8.5 billion this year, according to a report by The Information.

Microsoft has also struck a $650 million software licensing deal with Inflection AI (that is working on developing personalised AI models) and has acquired a minority stake in French startup Mistral AI for an investment amount of $16.3 million, according to TechCrunch.

Google

Google’s data centre investments have amounted to $3 billion, according to its second-quarter results. It has spent another $60 million to train its AI models on user posts on Reddit, Business Insider reported. Aside from this, the search giant has not disclosed concrete AI investment figures though company executives were confident that its AI strategy will yield long-term returns.

Google has been making steady progress with its AI models. In July, the company’s AI unit, DeepMind, unveiled new models that solved four of six mathematical problems included in this year’s International Math Olympiad. However, Google’s main focus is on bringing generative AI to its core search business which faces competitive risk from Microsoft’s Copilot-integrated Bing as well as new challengers such as PerplexityAI and SearchGPT.

Additionally, recent features such as AI overviews on Google Search and Gemini updates had to be rolled back and reevaluated due to inaccuracies.

Meta

On July 31, Meta reported its second-quarter earnings which revised the company’s capex estimates for the year from $35 billion-$40 billion to $37 billion-$40 billion. “We currently expect significant capital expenditures growth in 2025 as we invest to support our artificial intelligence research and product development efforts,” the company said.

Analysts had earlier projected that the tech giant’s expenses would hit $18 billion by 2024 and reach $50 billion by 2025, as per reports.

Thanks to these expenditures, Meta could have roughly 6,00,000 GPUs as CEO Mark Zuckerberg said that he is looking to buy over 3,50,000 Nvidia GPUs this year. Based on the earnings call, it seems that the company is mainly looking to set up computing clusters and data centres that could be needed for the AI future that lies ahead.

Meta’s large language model Llama 3.1 made its debut in July, and the next iteration of the open source AI model will need ten times more computing power than the current one, according to Zuckerberg. He also teased the company’s plans on developing AI tools specifically for advertisers. “AI is going to significantly evolve our services for advertisers in some exciting ways,” he said, adding that in the long term, “advertisers will basically just be able to tell us a business objective and a budget, and we’re going to go do the rest for them.”

Apple

Coming to Apple, the iPhone maker published its earnings for the third quarter on August 1.

Without specifically disclosing how much it has spent on AI, the Cupertino-based tech giant reported that its R&D spending came up to over $8 million in Q3 2024. Apple CEO Tim Cook further said that AI spending by the company will increase year over year.

“During the quarter, we were excited to announce incredible updates to our software platforms at our Worldwide Developers Conference, including Apple Intelligence, a breakthrough personal intelligence system that puts powerful, private generative AI models at the core of iPhone, iPad, and Mac,” Cook said.

“We very much look forward to sharing these tools with our users, and we continue to invest significantly in the innovations that will enrich our customers’ lives, while leading with the values that drive our work,” he added.

With Apple Intelligence, the company is mainly looking to leverage generative AI to transform user experience across its devices and thereby, boost sales. It also has a tie-up with OpenAI to integrate ChatGPT into its products, though the financial terms of the deal are still unclear.

Amazon

Announcing its financial results for the past quarter, Amazon said that it is looking to invest over $230 million in startups developing generative AI apps, out of which, $80 million will go towards funding the company’s AWS generative AI accelerator programme.

In March, Bloomberg reported that the e-commerce giant plans on spending around $150 billion constructing data centres over the next 15 years.

Amazon is pitching its AWS cloud services and infrastructure as the best place for app developers to build generative AI products by offering them compute credits. This could ensure that the company stays in the AI game as it reduces chances of developers shifting to other cloud service providers such as Microsoft’s Azure.

Notably, Amazon founder Jeff Bezos has invested in Perplexity AI which was recently valued at over $520 million.

As for its in-house efforts, Amazon is working on rolling out its own AI chatbot Metis to take on OpenAI’s ChatGPT. It is also looking to manufacture AI chips that could help the company cut down on buying chips from Nvidia to power its cloud business, according to a Reuters report.