ARTIFICIAL INTELLIGENCE · BIG TECH
Meta is making one of the technology industry's largest bets on artificial intelligence, investing billions of dollars in data centres, computing chips, AI researchers and new products. The company wants AI to strengthen Facebook, Instagram and WhatsApp while building more advanced assistants and wearable devices. But with spending rising sharply, Meta faces a major challenge: turning its enormous investment into lasting business value.
Meta, the company behind Facebook, Instagram, WhatsApp and Threads, is moving far beyond its traditional identity as a social media business.
Under CEO Mark Zuckerberg, the company is investing heavily in artificial intelligence, expanding its computing infrastructure, recruiting specialist researchers and developing new AI models. Its ambitions include improving the apps used by billions of people and creating more capable AI assistants that can help users perform tasks rather than simply answer questions.
The scale of the investment is striking. Meta's official financial guidance projected capital expenditure of between $130 billion and $145 billion for 2026, including principal payments on finance leases. The company said the spending reflects increased investment in infrastructure to support its AI efforts and core business.
That figure covers capital expenditure across the business, not AI alone. Nevertheless, AI computing capacity and data-centre expansion are major drivers of the increase.
Meta is effectively betting that artificial intelligence will become central to the future of its business, from advertising and content recommendations to personal assistants, business tools and smart glasses.
Why Meta Is Spending So Much on AI
Building advanced AI systems requires enormous computing resources. Companies need specialised chips, servers, data centres, electricity, cooling systems and highly skilled engineers to train and operate increasingly capable models.
Meta is investing in these foundations because it does not want to depend entirely on other companies to determine the future of AI.
The company's strategy includes developing its own AI models, expanding its data-centre capacity and building systems that can serve large numbers of users. More computing power can help researchers train advanced models and allow Meta to offer AI features across its products.
Meta also has a potential advantage that many AI startups do not: an established global audience and a large advertising business that can help finance its ambitions.
However, having billions of users does not automatically guarantee success in AI. Meta must still develop products people find useful and demonstrate that its infrastructure investments can generate sufficient economic returns.
Meta's AI Investment in Numbers
$130 billion–$145 billion
Meta's projected 2026 capital expenditure range in its July 2026 financial guidance.
$60.8 billion
Meta's reported revenue for the second quarter of 2026, up 28% year over year.
$31.08 billion
Capital expenditure, including principal payments on finance leases, reported for the second quarter of 2026.
3.60 billion
Average daily active people across Meta's family of applications during June 2026, according to the company's quarterly results.
These figures help explain why Meta can pursue AI projects on such a large scale. Its existing platforms generate substantial revenue, while its global reach gives the company many opportunities to introduce AI-powered services.
But the financial numbers also highlight the risk. Capital expenditure consumes cash today in the hope of generating benefits over many years. The returns are not guaranteed, particularly when technology changes quickly.
Source: Meta's second-quarter 2026 financial results.
Meta Wants to Compete With OpenAI and Google
Meta is competing in a market that includes OpenAI, Google, Anthropic and other companies developing advanced AI systems.
For years, Meta has developed its own large language models, including the Llama family. It has also integrated Meta AI into its applications and devices, giving users access to AI assistance within services they already use.
More recently, Meta has increased its focus on building advanced models through its Superintelligence Labs organisation.
In October 2026, reporting by the Financial Times described the launch of Meta's Muse Wonder model and Zuckerberg's renewed push to compete at the leading edge of AI development. The company is positioning its newer models and products as part of a broader strategy to deliver more capable AI experiences.
The important question is not simply whether Meta can release a powerful model. It must also make its systems reliable, useful, affordable to operate and attractive to users and businesses.
AI competition is becoming increasingly expensive. Companies must continually improve their models, secure computing resources and recruit people capable of advancing the technology. Falling behind could weaken their ability to attract customers and developers.
For Meta, building competitive AI is therefore both a product strategy and a long-term investment in the company's position in the technology industry.
Data Centres Are at the Centre of the Strategy
Much of Meta's AI investment goes into infrastructure that users never see directly.
Training and running advanced AI models requires large quantities of computing power. That means building or securing access to data centres filled with servers, networking equipment, storage systems and specialised chips.
Meta has been developing large data-centre projects in the United States and entering financing partnerships to support the expansion. In July 2026, the company announced a strategic venture with BlackRock to develop a data-centre campus in El Paso, Texas.
These facilities are intended to provide the computing capacity needed for AI development and the company's wider services.
Data centres also create practical challenges. They require electricity, water or other cooling arrangements, reliable networks and suitable locations. Delays in construction or shortages of equipment can slow expansion even when a company has the money to invest.
Meta's infrastructure strategy shows that the AI race is not only about writing better software. It is also about securing the physical resources needed to build and operate that software at scale.
Source: Meta's announcement on its BlackRock data-centre venture.
Meta Is Also Developing Its Own AI Chips
Advanced AI systems rely heavily on specialised chips. These processors perform the calculations required to train models and respond to users' requests.
Meta buys computing hardware from suppliers, but it is also developing its own chips through its Meta Training and Inference Accelerator programme.
Reuters reported in July 2026 that Meta planned to begin production of a new AI chip, code-named Iris, as part of its effort to expand computing capacity. The reported plans included a substantial increase in the company's available computing power over the following years.
Developing custom chips could give Meta more control over performance, energy efficiency and infrastructure costs. It could also help the company tailor its computing systems to the specific workloads generated by its AI products.
However, designing and manufacturing competitive chips is difficult. Meta will still need a dependable supply chain, advanced manufacturing partners and access to other essential components.
The goal is not necessarily to eliminate outside suppliers. It is to build a more flexible computing infrastructure that supports Meta's long-term AI ambitions.
How AI Could Improve Facebook and Instagram
Meta does not have to wait for a completely new AI business to generate benefits. Artificial intelligence can already improve the company's established services.
On Facebook and Instagram, AI systems help determine which posts, videos and recommendations appear in users' feeds. Better recommendation systems can help people discover more relevant content and spend more time using the applications.
AI can also improve advertising. By helping systems understand user interests and the context of content, Meta can potentially show more relevant advertisements and improve the likelihood that people interact with them.
That matters because advertising remains the company's principal source of revenue.
Meta's second-quarter 2026 results reported revenue of $60.8 billion, up 28% from the same quarter a year earlier. Advertising accounted for most of that revenue.
These results show that Meta's existing business remains a major financial foundation for its AI ambitions. They do not, by themselves, establish how much of the growth was caused by AI, because advertising demand, pricing, user activity and other factors also influence performance.
Over time, Meta hopes that AI will improve its products while opening opportunities for additional services and revenue streams.
The Next Goal: AI Assistants That Do More
Today's AI assistants can answer questions, summarise information, generate text and help users solve problems. The next stage of development is likely to involve systems that can carry out more complex tasks with less step-by-step instruction.
Meta has an opportunity to connect AI assistance to the services people already use, including WhatsApp, Instagram and Facebook. For businesses, this could mean tools that answer customer questions, help create marketing material or assist with routine communication.
For individual users, more capable assistants could eventually help organise information, plan activities and perform supported tasks across digital services.
However, the transition from answering a question to taking action introduces new risks. An AI system that sends messages, changes settings or handles business information needs appropriate permissions, safeguards and reliable ways to correct mistakes.
Meta's long-term opportunity depends not just on making its assistants more powerful, but on making them dependable enough for people and businesses to trust.
Smart Glasses Could Become Another AI Platform
Meta is also investing in wearable technology, including smart glasses developed with eyewear partners such as Ray-Ban and Oakley.
These devices can combine cameras, microphones, speakers and AI assistance in a form that people can wear throughout the day. Depending on the model and supported features, users can take photos, record videos, listen to audio and ask questions through voice commands.
Smart glasses give Meta a way to bring AI outside the smartphone screen. A wearer could potentially ask for information about something in front of them or access assistance while keeping their hands free.
That creates opportunities in accessibility, travel, communication and content creation. But it also raises important questions about privacy, recording consent, battery life, comfort and the accuracy of AI-generated information.
Meta's investment in glasses suggests that the company sees AI as more than a chatbot inside an application. It wants its technology to become part of how people interact with the physical world.
The Financial Risk Behind Meta's AI Ambitions
Spending billions on AI does not guarantee that a company will dominate the market.
Meta's projected 2026 capital expenditure of $130 billion to $145 billion represents an enormous commitment. The company must pay for infrastructure and technical resources long before it can know the full financial return on every investment.
Investors are therefore watching whether AI produces enough additional revenue, improves operating efficiency or creates new business opportunities to justify the spending.
There are several risks.
- High infrastructure costs: Data centres, chips, electricity and cooling systems are expensive to build and operate.
- Rapid technological change: Today's advanced hardware and models may become less competitive as new approaches emerge.
- Uncertain revenue: Users may enjoy AI features without paying directly for them, making monetisation more difficult.
- Competition: OpenAI, Google, Anthropic and other companies are also investing heavily in research and computing capacity.
- Regulation and privacy: AI products may face restrictions involving personal data, copyright, consumer protection and safety.
Investors also need to consider the effect of higher spending on cash flow and profitability. Meta's strong advertising business gives it a substantial source of funding, but the scale of its investment makes financial discipline increasingly important.
The company has argued that AI can strengthen its core business and support future products. Whether the full investment pays off will become clearer as its products develop and financial results reveal more about their contribution.
What Meta's AI Push Means for the Future
Meta's investment reflects a broader shift across the technology industry. Large companies are treating AI as a foundational technology that could influence search, advertising, communication, software, shopping and personal computing.
If AI systems become more capable and affordable, they could change how people work and how businesses serve customers. Companies that control the underlying infrastructure and develop widely used products may gain significant advantages.
But there is another side to the transition. The concentration of computing resources in a small number of technology companies could increase their influence over digital services. AI systems also raise concerns about privacy, misinformation, employment disruption and the reliability of automated decisions.
For Meta, the challenge is to turn its investment into products that people genuinely value while managing the financial, technical and social consequences of deploying AI at a global scale.
The company is not simply adding another feature to Facebook. It is attempting to position itself for a future in which AI plays a much larger role in how people communicate, discover information and use digital tools.
Key Facts
- Company: Meta Platforms, the parent company of Facebook, Instagram, WhatsApp and Threads.
- 2026 capital expenditure outlook: $130 billion to $145 billion in the company's July guidance.
- AI priorities: Advanced AI models, data centres, computing chips, AI assistants and wearable devices.
- Quarterly revenue: $60.8 billion in Q2 2026, up 28% year over year.
- Infrastructure strategy: Expanding computing capacity through data-centre projects and partnerships.
- Main challenge: Generating sustainable financial returns from exceptionally large investments.
Meta is making a high-stakes bet on artificial intelligence. Its existing social platforms provide revenue and access to billions of people, while its investments in models, chips, data centres and smart glasses could help it expand into new areas of technology.
The real test will be whether those investments produce lasting advantages and useful products. In the AI race, spending at scale can create opportunities, but the companies that succeed will also need reliable technology, strong execution and a sustainable business model.
