OpenAI made an acquisition last month that should make every competitor—especially Apple—sweat: they bought Software Applications Inc., the tiny startup behind an AI assistant called Sky. The team they acquired? The same genius developers who built Apple's Shortcuts app.
This is the clearest signal yet that OpenAI is moving beyond the "chatbot" phase and into the "autonomous agent" phase. They are building the infrastructure to make ChatGPT do things instead of just talk about them.
Beyond the Sandbox
OpenAI has been trying to make ChatGPT "agentic" for a while, launching a web browser (Atlas) and building integrations. But those are all sandboxed experiences with limited scope.
Sky changes the game because it gives ChatGPT native access to macOS. Sky operates using macOS accessibility APIs, allowing it to see your screen, understand context across multiple windows, and operate any app just like a human would.
Early testers described it as transformative. You could tell Sky, "Turn this webpage into a calendar event and message my team," and it would parse the content, create the event, and send the message across different apps without you touching anything. That’s a fundamentally different level of integration than copy-pasting an email draft.
The Apple Problem
This move puts OpenAI in direct, accelerated competition with Apple. Apple is currently overhauling Siri with Apple Intelligence to do exactly what Sky does—understand on-screen content and perform actions—but that overhaul isn't expected until spring 2026.
OpenAI just bought the people who know macOS automation better than almost anyone outside of Cupertino, and they are integrating that expertise into a product (ChatGPT) that already has hundreds of millions of users. Apple now has a major external threat moving deeper into the OS layer while their internal strategy plays catch-up.
The Privacy Conversation Nobody is Having
For Sky's technology to work, it needs to see everything on your screen to understand context—your emails, documents, browser tabs, and banking data.
Apple built its brand on privacy, using on-device processing or Private Cloud Compute for AI. Sky's technology probably runs in the cloud, meaning your screen content is being processed on OpenAI's servers. OpenAI hasn't detailed how they will handle this enormous privacy risk, but any system that can "see" your screen is inherently high-risk if compromised.
My Take
This acquisition is the ultimate strategic power grab. OpenAI saw an opening to be first to market with native OS-level AI automation on Mac, and they took it.
The technology is exciting, but the safety and privacy implications are terrifying. An AI that can see your screen and take actions sounds amazing until it does the wrong thing at the wrong time. If OpenAI doesn't absolutely nail the UX of user control and security, this could become more intrusive than helpful. The technical battle is over; the design and trust battle is just beginning.