AI labs race to buy Mac minis for smarter computer agents
AI labs aren’t just hoarding GPUs anymore—they’re clearing out Apple store shelves. OpenAI has purchased tens of thousands of Mac minis and Mac Studios to train computer-use agents, according to a report by The Information. The demand is so fierce that even Apple’s most powerful desktop models have been sold out for months, pushing the company’s Mac revenue up 29 percent to $10.4 billion in the June quarter.
Why the sudden hunger for Cupertino’s compact machines? Computer-use agents—AI systems designed to navigate software, automate tasks, and interact with interfaces the way a human would—need reliable, standardized hardware to train efficiently. Mac minis, with their consistent macOS environment and Apple Silicon chips, offer a stable platform for collecting high-fidelity interaction data without the variability of cloud-based setups. Anthropic is also reportedly relying on Apple hardware for similar purposes, signaling a broader industry shift toward diversified training infrastructures.
The move underscores a quiet arms race in AI development: as models grow more capable of manipulating user interfaces, the hardware beneath them must keep pace. Apple’s supply constraints hint at a supply chain strain that could ripple across the tech ecosystem, affecting not just AI labs but also developers building on macOS. For now, the Mac mini’s affordability and efficiency make it an attractive stopgap, but the long-term implications for hardware diversity in AI training remain unclear.
Why it matters
This isn’t just about hardware shortages—it’s a glimpse into how AI’s next frontier depends on the mundane tools we use daily. By locking down Apple’s compact desktops, labs are betting on consistency over raw power, a strategy that could reshape how we think about training environments. For developers and businesses, the trend highlights a growing reliance on consumer-grade hardware, raising questions about accessibility and supply chain resilience in the AI supply chain. The stakes extend beyond silicon: they define the very infrastructure of tomorrow’s intelligent systems.
Source: The Decoder. AI-assisted editorial synthesis — TechnoExpress.