OpenAI Is Reportedly Developing a Phone That Makes Apps Obsolete

Apps as we know them could soon be a thing of the past. At least, that is what the latest report from analyst Ming-Chi Kuo, one of the most closely watched supply-chain experts in the technology industry, suggests. According to his post on X, OpenAI is working on its own smartphone in collaboration with chipmakers Qualcomm and MediaTek and manufacturing partner Luxshare. Mass production is expected to begin in 2028.

None of the companies involved—OpenAI, Qualcomm, or MediaTek—has confirmed the report. Nevertheless, Qualcomm’s stock jumped 13 percent before trading even began.

The end of apps. Really?

Kuo is not describing a phone with an improved voice assistant. He is describing a device where an AI agent replaces the entire interface. No icons. No switching between apps. Instead, there is a single intelligent layer that you tell what you want, and it gets it done. Want to book a flight? Order food? Send a message? The agent handles it itself, without a single tap on an icon. “Users aren’t trying to use a bunch of apps,” Kuo wrote. “They’re trying to complete tasks and meet needs using their phone. This fundamentally changes the way people think about smartphones.”

The phone would operate on two layers. Lighter tasks, such as tracking context, managing memory, or running smaller AI models, would take place directly on the device. More complex computations would be sent to the cloud. According to Kuo, the device would maintain a “full real-time state” and continuously monitor the user’s location, activity, communications, and surroundings.

Today, OpenAI operates as an app on phones made by Apple or Google. That comes with major limitations. Both companies determine what level of system access apps receive and what they are allowed to do. That is not enough for a true AI agent capable of autonomously controlling the entire device.

Kuo argues that only by having full control over both the operating system and the hardware can OpenAI deliver a complete service powered by AI agents. Its own phone would also provide direct access to user behavior data, more detailed than anything a third-party app can obtain. ChatGPT is approaching one billion weekly users. A device designed for everyday use could multiply that reach in a way no app can.

Who will make such a phone?

Supply-chain company Luxshare assembles AirPods and parts of the Apple Watch and is taking over an increasingly large share of iPhone production. Qualcomm supplies Snapdragon 8 Elite Gen 5 chips for 75 percent of Samsung Galaxy S26 phones. MediaTek matches Snapdragon in performance and surpasses it in power efficiency. These are not suppliers for a concept. They are suppliers for phones that sell hundreds of millions of units annually.

The phone’s specifications and supplier list are expected to be finalized by the end of 2026 or in the first quarter of 2027. Qualcomm also has a strategic interest of its own. Before the report was published, its shares were trading at $149.84, well below last year’s high of $205.95. A partnership with OpenAI would open up a new source of revenue at a time when Apple is developing its own modems and MediaTek is pushing into the premium market segment.

Two projects at once

The phone is not the only hardware project OpenAI is working on. In parallel, it is collaborating with former Apple chief designer Jony Ive, whose company io is developing a smart device without a display. According to reports so far, the first product will be a smart speaker with a camera, followed by glasses, a lamp, and headphones. The first product in this line is expected to arrive in early 2027.

Meanwhile, OpenAI CEO Sam Altman published a telling post on X: “I think it’s time to seriously rethink how operating systems and user interfaces are designed.”

An OpenAI smartphone would not be the first attempt to have artificial intelligence replace the phone or apps. The $699 Humane AI Pin came to an end in February 2025, when HP bought the remnants of the company for $116 million and shut down its servers. The $199 Rabbit R1 received 100,000 preorders, but after five months only 5,000 people were actively using it. The founder himself admitted that the product had launched too early.

Both projects failed for the same reason: they added a new device to the pocket of a user who already had a phone there. OpenAI is betting on a different approach. It does not want to add. It wants to replace. Kuo estimates annual sales of 300 to 400 million units, which would put the phone ahead of both the iPhone and Samsung in terms of volume. Apple sells roughly 230 million iPhones annually, while Samsung sells approximately 220 million Galaxy phones. No new manufacturer in the smartphone era has ever reached such figures.

By 2028, OpenAI must finalize the specifications, secure component supplies, build manufacturing capacity, create an agent-based software platform, arrange distribution through mobile carriers, launch retail sales, and convince hundreds of millions of people to replace their iPhone or Galaxy with a phone from a company that has yet to manufacture a single piece of consumer electronics.

Sources: cnet.com and 9to5mac.com