Imagine picking up your phone and never opening an app again. No grid of icons, no app store, no toggling between tabs. You just tell the device what you need, and an AI agent handles the rest.
That is the vision behind a smartphone concept reportedly in development at OpenAI, according to a late April 2026 note from supply-chain analyst Ming-Chi Kuo. The device, which Kuo says targets mass production around 2028, would ditch the conventional app model entirely in favor of AI agents that book flights, manage calendars, shop, and negotiate customer-service calls on a user’s behalf. Bloomberg reported that Qualcomm shares soared in the wake of Kuo’s note.
The timing is notable. Just weeks earlier, OpenAI closed a $6.5 billion deal to absorb io Products, the hardware company led by Jony Ive, the designer behind the iPhone, iPad, and MacBook. If the phone project moves forward as described, it would pit OpenAI directly against the app-store empires that Apple and Google have spent nearly two decades building.
What the supply-chain report actually says
Kuo, who built his reputation on accurate Apple supply-chain predictions, shared his findings on X. He wrote that OpenAI’s phone would use processors from both MediaTek and Qualcomm, with Chinese manufacturer Luxshare serving as the exclusive system co-design and manufacturing partner. The device, in his words, is meant to “deliver a comprehensive AI agent service” rather than rely on downloadable applications.
That phrase points to a fundamentally different interaction model. Instead of hunting for a flight-booking app or a restaurant-reservation app, a user would describe a task in plain language and an AI agent would carry it out across multiple services behind the scenes.
Kuo’s track record gives his notes outsized influence on tech stocks. For chipmakers and contract manufacturers, simply being named in one of his reports can shift expectations about future orders, which helps explain why Qualcomm jumped so quickly. But analyst intelligence is not a signed contract, and Kuo himself has occasionally been early or wrong on specifics even when the broad direction proved correct. It is worth stressing that the partnership details about MediaTek, Qualcomm, and Luxshare originate from Kuo’s supply-chain note, not from any official company announcement. Bloomberg’s reporting covered the resulting stock reaction, not the underlying partnership claims themselves.
What OpenAI has confirmed
OpenAI’s hardware ambitions are not speculation. A joint letter from CEO Sam Altman and Ive, published on OpenAI’s website, announced the io Products merger and described a vision of devices that move beyond screens and existing computing conventions. The letter does not name a specific phone product, reference AI agents replacing apps, or mention any of the supply-chain partners in Kuo’s note.
According to that letter, the broader collaboration between Altman and Ive’s design firm LoveFrom began around 2023, initially as exploratory work that predated a formal company. io Products itself was founded around 2024 and later merged into OpenAI. In other words, the 2023 date refers to the early partnership between the individuals, while the 2024 date marks the creation of the corporate entity that eventually became part of OpenAI. That timeline suggests the hardware effort predates Kuo’s report by years and is not a reaction to the latest wave of AI hype. It appears to be a deliberate, multi-year strategy to pair OpenAI’s software with purpose-built devices.
The $6.5 billion price tag reinforces the seriousness. Companies do not spend that kind of money on side experiments. Whatever OpenAI builds, the financial commitment and design talent are real.
What remains unconfirmed
No one at OpenAI, MediaTek, Qualcomm, or Luxshare has publicly confirmed the specific phone project Kuo describes. His claims rest on supply-chain intelligence, not official announcements or leaked prototypes. The 2028 mass-production target is a projection, not a committed date, and consumer electronics timelines frequently slip by a year or more during early development.
The relationship between the Ive merger and Kuo’s phone report also lacks a direct, on-the-record connection. It is possible that Ive’s team is working on an entirely different device category, such as a wearable or a home product, or that the phone project and the Ive collaboration are parallel tracks that have not yet converged publicly.
Conceptual illustrations of the phone’s interface have appeared in TechCrunch’s coverage, but these are described as mock-ups, not leaked product images. No technical specifications, screen sizes, battery details, or pricing have surfaced from any source. The entire product description, at this stage, is built on Kuo’s analyst note and OpenAI’s general statements about hardware intent.
The ghost of AI hardware past
OpenAI would not be the first company to promise a phone that replaces apps with AI. The Humane Ai Pin and the Rabbit R1 both launched with similar rhetoric and landed to brutal reviews. The Ai Pin was criticized for sluggish performance, overheating, and a subscription model that delivered less than a basic smartphone. The Rabbit R1 struggled with reliability and a narrow set of supported services. Both devices demonstrated that the “just talk to it” concept is far easier to pitch than to execute at the speed and accuracy consumers expect.
OpenAI has advantages those startups did not: a leading large language model, billions in capital, a world-class hardware designer, and relationships with major chipmakers. But the failures of earlier AI-first devices highlight the gap between a compelling demo and a product people actually use every day. Reliable access to third-party services, consistent handling of privacy and data permissions, and robust safeguards against errors are all unsolved problems at consumer scale.
Why this would threaten Apple and Google
If the project proceeds as Kuo describes, the most disruptive element would not be the industrial design or the processor choice. It would be the software model. Today’s phones revolve around app stores, icons, and user-initiated taps. An AI-first device would invert that logic: the primary interface would be a conversational agent that orchestrates services on the user’s behalf.
For consumers, that could reduce friction and make complex tasks more accessible, especially for people who struggle with dense menus or fragmented app ecosystems. For developers and service providers, it could upend distribution and discovery, shifting attention away from app rankings and toward how well their services integrate with dominant AI agents.
For Apple and Google, whose mobile businesses depend heavily on app-store control and in-app purchase fees, a successful AI-native phone would present a direct competitive threat. It is worth noting, though, that OpenAI already partners with both companies. ChatGPT is integrated into Apple’s Siri, and OpenAI’s models power features across Samsung’s Galaxy devices. Building a rival phone would complicate those relationships considerably.
Regulators may also have questions. How would an AI agent decide which airline to book or which restaurant to recommend? Would it favor partners who pay for placement? None of those questions are addressed in current public documents about OpenAI’s hardware plans.
Signals to watch through mid-2026
Without official confirmation, the most telling signals will come from hiring patterns, regulatory filings, and public statements from the companies Kuo named. Large-scale smartphone projects leave traces in component orders and manufacturing capacity plans long before launch. If OpenAI intends to ship a mass-market device around 2028, deeper supply-chain reporting and more concrete leaks should emerge well before that date.
For now, the OpenAI phone is a credible but unproven idea: backed by a marquee designer, significant capital, and a respected analyst’s supply-chain intelligence, yet still confined to projections and high-level corporate vision statements. The real story is not about a guaranteed product. It is about where one of the most influential AI companies is pointing its resources, and whether it can succeed where every previous attempt to kill the app has failed.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
