Meta is putting a much bigger business bet behind artificial intelligence.
On August 10, 2026, Mark Zuckerberg outlined a vision in which advanced AI becomes a personalised tool available to individuals rather than technology controlled by a small group of companies or governments.
The timing matters. Meta generated $60.8 billion in revenue in the second quarter, but its costs rose 55%, while its 2026 capital expenditure guidance has reached as much as $145 billion.
Open AI Takes Centre Stage
Zuckerberg’s latest argument is closely tied to Meta’s renewed push for open-weight AI models. Alongside his essay, Meta launched Muse Glimmer, a model designed for smaller agentic tasks on personal devices and capable of running using a single graphics card, according to Reuters.
Meta also said it plans to release Muse Spark 1.2, which it describes as its most advanced model yet.
The strategy marks a practical extension of Meta’s earlier open-model approach. Instead of keeping every advanced model behind a closed service, Meta is making model weights available in at least some cases, allowing developers greater freedom to adapt and deploy them.
That approach also fits Meta’s wider product strategy. The company has already positioned Muse Spark as a core model for Meta AI across its consumer ecosystem.
Meta introduced Muse Spark in April as the first model from Meta Superintelligence Labs and said it would power Meta AI across its app and website before expanding into other products.
Spending Raises Stakes
The financial backdrop makes the strategy more consequential.
Meta’s Q2 2026 revenue rose 28% year-on-year to $60.801 billion. Advertising remained overwhelmingly responsible for that business, generating $59.363 billion during the quarter.
At the same time, total costs and expenses increased 55% to $42.026 billion. Net income declined 14% to $15.848 billion, while operating margin dropped to 31% from 43% a year earlier.
Meta also spent $31.08 billion on capital expenditure during the quarter. For the full year, it expects capital expenditure of $130 billion to $145 billion.
Importantly, this is Meta’s total capex guidance and should not be treated as an AI-only budget. The company says the increase reflects investments in infrastructure supporting its AI ambitions as well as its broader business.
That distinction matters because the headline figure can otherwise exaggerate the amount Meta is directly allocating to AI.
Infrastructure Is Strategic
Meta’s AI ambitions extend well beyond models. In March, the company said it was developing and deploying four new generations of its Meta Training and Inference Accelerator, or MTIA, chips within two years. Meta said hundreds of thousands of these chips were already being deployed for inference workloads across its platforms.
The company has also described MTIA as central to its AI infrastructure strategy, particularly for recommendation systems, advertising-related workloads and generative AI inference. That gives Meta more control over the computing infrastructure required to serve AI at its enormous user scale.
The scale of that user base is important. Meta reported 3.60 billion daily active people across its Family of Apps in June 2026, up 3% from a year earlier.
For Meta, therefore, AI does not need to become a standalone consumer business overnight to have commercial significance. It can first influence products that already reach billions of people.
Meta’s Bigger AI Bet
Zuckerberg’s personal-superintelligence vision is therefore both a technology strategy and a capital-allocation decision.
Meta has the distribution, advertising business and infrastructure resources to pursue the idea at a scale few companies can match. Its 3.60 billion daily users provide an unusually large potential audience, while its investment in custom chips and computing infrastructure is designed to support AI workloads across the company.
The harder question is whether that scale can justify the cost.
With 2026 capital expenditure projected at up to $145 billion, Meta is committing substantial resources to infrastructure while operating margins have already come under pressure.
For now, Zuckerberg’s promise is clear: advanced AI should become personal, widely accessible and increasingly open. The business test will be equally clear. Meta must show that the enormous infrastructure and model investments behind that vision can strengthen the products and advertising business that ultimately finances the bet.
Also Read: JPSC-JSSC Protest Turns Tense In Ranchi; BJP Calls Jharkhand Bandh After Police Action
