OpenAI and other major AI labs have purchased tens of thousands of Apple Mac minis and Mac Studios to train computer-use agents, according to The Information. This unexpected surge in demand reflects a new frontier in AI development: teaching AI systems to interact with computers the way humans do.
Computer-use agents represent a shift in AI capabilities. Rather than processing text or images, these systems learn to click buttons, type text, navigate interfaces, and execute tasks across desktop environments. Training these agents requires running them on actual computers or simulated desktop systems. Mac hardware provides a stable, controllable platform for this work.
Anthropic, another leading AI lab, similarly depends on Apple hardware for this training process. The demand has proven so intense that Apple's most powerful Mac models have experienced persistent stockouts lasting months. This supply constraint underscores how seriously these companies treat computer-use agent development.
Apple's financial results validate the scale of these purchases. Mac revenue jumped nearly 29 percent to $10.4 billion in Apple's June quarter. This growth reflects both consumer demand and the wholesale bulk purchases from AI labs. For context, this marks one of Mac's strongest performances in years, driven largely by institutional buyers rather than individual consumers.
The rationale for choosing Mac hardware over alternatives makes sense. First, consistency matters for AI training. Mac environments are uniform and standardized, reducing variability in training data. Second, these systems need reliability. Crashes or inconsistencies during months-long training runs waste computation and delay progress. Third, macOS offers Unix-based stability with tools that developers already use.
The shift also highlights a strategic divergence in AI development. OpenAI's focus on computer-use agents complements its broader strategy to create AI that operates autonomously across software ecosystems. Anthropic's parallel investment suggests this capability will define the next generation of competitive AI. Both companies recognize that systems capable of using computers independently unlock new applications in software automation, research, and business processes.
The Mac mini purchase strategy reveals pragmatism. Mac Studios cost significantly more than Mac minis but offer additional computing power. Buying tens of thousands of Mac minis balances cost and capability. Each unit still provides substantial CPU performance for training, and the total spending remains manageable compared to custom AI hardware. This approach differs from the traditional path of building specialized training infrastructure.
Supply chain implications extend beyond Apple. The sustained demand for Mac hardware strains Apple's production capacity and inventory. Component suppliers see increased orders. Shipping and logistics networks feel pressure. This cascading effect demonstrates how AI development cycles now directly influence consumer hardware markets.
For Apple, this represents an unexpected revenue stream and validation of Mac architecture's general-purpose computing capabilities. The company gains leverage in enterprise AI discussions, though it remains primarily a hardware supplier rather than a primary AI developer. Apple's own AI initiatives have focused on on-device processing, creating an interesting dynamic where Apple's hardware enables competitors' cloud-based AI training.
The trend will likely continue. As computer-use agents mature and prove valuable, demand for training infrastructure grows. Other AI companies may follow OpenAI and Anthropic into Mac procurement. Apple may eventually need to expand production or create specialized configurations for AI labs. This market segment, currently nascent, could evolve into a meaningful component of enterprise hardware sales.