OpenAI is moving into hardware, and it is aiming straight at the device layer Apple has owned for two decades. The company is reportedly on track to unveil its first consumer device in the second half of 2026 — a screenless, always-listening companion built on the $6.5 billion acquisition of io Products, the studio founded by former Apple design chief Jony Ive. Early reports describe a smart-speaker-style product priced around $200 to $300 with an onboard camera and environmental sensors, the first of a planned family of context-aware devices designed to make ChatGPT-class AI ambient rather than something you open on a phone.
Apple, for its part, is said to be developing an AI-powered wearable pin that may not arrive until 2027 — a notable gap in a market where earlier entrants like Humane failed precisely because the hardware outran the AI. The strategic stakes are larger than gadget sales. If interactions move from tapping glass to talking to an always-on assistant, the company that owns the device also owns the default model, the context and the data. That is the same control point Apple has defended with the iPhone, and OpenAI is now contesting it directly rather than living as an app inside someone else's ecosystem.
For Singapore businesses the more immediate issue is governance, not gadget envy. Always-listening, camera-equipped AI devices in meeting rooms, factory floors and executive offices create real exposure under the PDPA and most corporate confidentiality policies. IT and security teams should decide now where ambient AI hardware is permitted, how recordings and inferences are stored and where they are processed, and whether data leaves Singapore. Voice and ambient computing will land in the enterprise whether or not it is sanctioned — the firms that write a clear acceptable-use and data-residency policy ahead of launch will avoid scrambling after the first device walks through reception.
