On Thursday, OpenAI announced its acquisition of a software startup behind Sky—a Mac application that enables users to control their computers through natural language commands. The entire Sky team will join OpenAI to integrate the app’s capabilities directly into ChatGPT.

This move marks OpenAI’s clearest step yet toward native desktop integration. While ChatGPT currently operates within browsers and standalone apps, Sky is engineered to float above the desktop and interact directly with the operating system—an approach that emphasizes ambient assistance over destination-based experiences.

The Sky acquisition follows just two days after OpenAI launched ChatGPT Atlas, an AI-powered browser built specifically for Mac. This timing is no coincidence but part of a deliberate strategy. Atlas positions ChatGPT at the core of web browsing with an agent mode capable of executing multi-step tasks. Sky extends that same logic to the OS level, empowering ChatGPT to observe and manipulate native Mac applications. Together, they form OpenAI’s blueprint for dominating the entire computing interface on Apple platforms.

Competitors are racing in the same direction: Anthropic’s Claude has been advancing computer-use capabilities, Microsoft has deeply embedded Copilot into Windows, and Google is experimenting with agent-like features in Search and Assistant. Although no company has yet nailed the ideal interface for AI to perform real-world tasks on computers, the race to define this space is intensifying.

OpenAI’s messaging echoes a familiar promise: “AI that understands context and adapts to your intent.” The tougher question remains whether users truly want an ever-present AI layer on their desktops. While existing Sky users clearly see its value, broad consumer demand for always-on AI interfaces remains unproven at scale.