OpenAI's global affairs head, Chris Lehane, told attendees at the Davos forum that the company is pursuing plans to launch its first hardware product in the second half of 2026. He stopped short of a firm promise, saying the latter half of 2026 is "the most likely" window but that the timetable depends on how development proceeds. Lehane offered no details on form factor, functionality or distribution, and said more information would follow later in the year.
The remarks fit a narrative OpenAI has been cultivating since its acquisition last year of io, the design studio founded by former Apple design chief Jony Ive. CEO Sam Altman has long suggested OpenAI intends to ship a device that embeds its AI more directly into users' daily lives, and promotional material from io hinted at public reveal in 2026. Media reports have added colour: the company is said to be experimenting with compact, screenless prototypes and a high-priority audio product, codenamed Sweetpea, that could plausibly compete with earbuds such as AirPods.
A hardware line would represent a strategic pivot for OpenAI from a business model dominated by subscriptions and compute revenues. The company’s CFO, Sarah Friar, reported an annualized 2025 revenue figure north of $20 billion — a sharp step up from about $6 billion in 2024 — driven tightly by increased spend on compute. Still, OpenAI has been exploring additional revenue channels, including advertising inside ChatGPT for some U.S. users, and hardware sales offer both direct revenue and a lever to tie users more tightly to its services.
Product and engineering challenges are substantial. If the device is indeed a small wearable or audio product, OpenAI must reconcile latency, battery life and privacy concerns with the computational appetite of modern large models. That tension points to likely hybrid architectures combining local inference for low-latency features and cloud-based compute for heavier tasks, or the use of highly optimized on-device models. Altman has described the intended experience as "more peaceful" than a smartphone, implying a pared-down interface and a different interaction model that emphasises ambient assistance rather than screen-based multitasking.
Market dynamics will matter as much as technology. The audio and wearable markets are crowded with incumbents such as Apple, Google and Meta, all of whom have deep hardware, OS and supply-chain expertise. Reports that OpenAI might target tens of millions of units in an initial year, if accurate, point to ambitious manufacturing and distribution plans and to a bet that the brand and software advantage can overcome hardware incumbency. Delivery at that scale would also expose OpenAI to component shortages, margin pressure and logistical complexity uncommon for a software-first company.
Beyond product and profit, a move into hardware carries regulatory and strategic consequences. A device that routes voice, biometric or contextual data through OpenAI systems will attract scrutiny over privacy, data flows and jurisdictional controls, especially amid rising global attention to AI governance. Governments and competitors will watch whether OpenAI’s device deepens its ecosystem and expands its influence over user attention and data in ways that could invite antitrust or security reviews. For investors, customers and rivals, the timetable and design choices Lehane characterised as contingent are signals that OpenAI is accelerating toward an integrated hardware-software future — but one that still faces technical, commercial and political headwinds.