OpenAI develops AI agent smartphone with Qualcomm and MediaTek, aims for 300-400 million annual shipments by 2028

According to TF International Securities analyst Ming-Chi Kuo, Qualcomm and MediaTek are co-designing a custom processor and Luxshare Precision Industry is developing a smartphone built around artificial intelligence agents rather than apps. If the device succeeds, Kuo predicts annual shipments of 300-400 million, surpassing Apple’s iPhone volumes and putting the phone in direct competition with two companies that control about 40% of the global smartphone market. Specifications and a list of suppliers are expected to be finalized in late 2026 or the first quarter of 2027, with mass production targeted for 2028. Shares of Qualcomm rose as much as 13% in premarket trading on the report. None of the three companies, Qualcomm, OpenAI or MediaTek, have confirmed the partnership. This is an analytical report, not an announcement, but the supply chain described by Kuo is not speculative. It’s the supply chain that builds the devices you already own.

Concept

The phone described by Kuo is not a smartphone with an AI assistant. It is a device where the AI ​​agent is the interface and the application is obsolete. Instead of downloading apps and navigating screens, users will interact directly with agents who perform tasks: book transportation, book restaurants, manage email, conduct research, text. The architecture will process lighter tasks on the device, including context awareness, memory management, and smaller AI models, while performing complex inference in the cloud. The device will maintain what Kuo calls “full real-time state,” continuously capturing the user’s location, activity, communication and environmental context to feed agents. That’s Qualcomm CEO Cristiano Amon’s vision for 2026: AI agents will replace the mobile operating system and apps as the primary interaction layer, and hardware must be designed from the ground up to support sustainable, energy-efficient AI results, rather than upgrading existing chipsets with neural processing units.

The concept is separate from OpenAI’s other hardware project, a smart speaker with a camera first, then glasses, a light bulb, and headphones, with former Apple design chief Jony Ive developing the company’s io phone-less device, the first product expected in early 2027. OpenAI follows two parallel hardware strategies: a device without a screen and a device that looks like a PC. it keeps the phone’s form factor but replaces everything that runs on it. Apple is testing smart glasses powered by artificial intelligence with a special chip, cameras and Siri equipped with a Gemini model targeting 2027. The question of whether artificial intelligence will live on your phone, on your face, or in the speaker on your counter is being answered by every major tech company at the same time. OpenAI is betting on it all at once.

Supply chain

The validity of the report depends on the supply chain, not the concept. Luxshare Precision Industry is a major Apple supplier that assembles a growing share of AirPods, Apple Watch components and iPhones. Qualcomm’s Snapdragon 8 Elite Gen 5 powers 75% of Samsung’s Galaxy S26 lineup, beating Apple in multi-core and GPU performance for the first time. MediaTek’s Dimensity 9500 matches Qualcomm and Apple in CPU performance at a lower price with better efficiency. These are not concept phone suppliers. They are the suppliers of hundreds of millions of phones sold. Qualcomm’s acquisition of Edge Impulseleading AI developer platform, signaled the company’s strategic commitment to achieve on-device artificial intelligence across device categories by 2025. The Snapdragon 8 Elite Gen 5 delivers 37% faster AI processing than its predecessor’s Hexagon NPU, supports agent AI that learns from user behavior, and includes a personal knowledge graph and continuous context awareness through an improved sensing hub. Qualcomm is also building custom 3D DRAM specifically optimized for AI workloads in mobile devices. Silicon doesn’t need to be invented for the phone Kuo describes. Components are available. The question is whether the software paradigm is viable.

Financial context is important. Shares of Qualcomm were trading at $149.84 ahead of the report, down from a 52-week high of $205.95, with profit growth of 46.9% and gross margin down 55.1%. The company reports earnings on April 29, two days after Kuo’s report. In February, Bloomberg reported that Qualcomm issued a “weak forecast in a sign of phone market turmoil.” The OpenAI partnership will represent a new revenue stream in a market where Qualcomm’s traditional business of supplying phone makers with modems and processors is under pressure from Apple’s efforts to develop its own modem chips and MediaTek’s incursion into the premium Android segment. Qualcomm will help build a device designed to challenge the iPhone while continuing to supply Apple with modem chips until at least 2027, a business relationship that epitomizes the contradictions of the semiconductor supply chain.

Cemetery

The AI ​​device category produced more failures than products. The $699 Humane AI Pin, a wearable with a laser projector that transmits data to the user’s palm, was bricked for good on February 28, 2025, when HP acquired the remains of Humane for $116 million and shut down the servers. The Rabbit R1, a $199 “large action model” device, attracted 100,000 pre-orders but retained only 5,000 active users after five months, a 95% churn rate. Its founder admitted that the device was released too early. Both failed for the same reason: they created new form factors that didn’t solve any problems the smartphone didn’t, at price points that required the user to carry a second device. The OpenAI phone takes a fundamentally different approach. Not an additional device. It replaces a device that 4.7 billion people already carry in the same form factor, with the same basic capabilities but with a fundamentally different interaction model. Whether that’s enough to avoid the graveyard depends on whether agents can do what apps do better, faster, and without the hassle of learning a new paradigm.

Artificial intelligence is already reshaping the mobile app ecosystem“vibe-coded” apps flood the App Store in such volume that Apple has been forced to crack down on submissions. The EU is set to force Google to open up Android to rival AI assistants Voice activation and deep integration, including ChatGPT and Claude, require equal system-level access under the Digital Markets Act. The software layer of the smartphone is already in motion. Samsung’s Galaxy S26 is powered by a triple AI engine with Gemini, Perplexity and Bixby. Google’s Pixel 10 hands off multi-step tasks to AI agents in the background. Apple Intelligence processes requests on the device with privacy in mind. Every major phone maker is moving headfirst into AI experiences, but they’re all limited by backwards compatibility with the billions of existing apps and operating systems that run them. The advantage of OpenAI is that it has no legacy if the phone becomes a reality. It can design a simple interaction model without worrying about whether Instagram’s notification system is working or whether the banking app is displaying correctly. The disadvantage is that users may not want a clean page. They may want their own apps and an AI assistant that works around them, which is what Samsung, Google and Apple already offer.

question

Kuo’s forecast of 300-400 million annual shipments would make the OpenAI phone one of the most successful consumer electronics products in history. For context, Apple ships about 230 million iPhones a year. Samsung ships about 220 million Galaxy phones. A new entrant reaching these volumes is unprecedented in the smartphone era. The prediction reflects the scale of OpenAI’s ambition, not a reasonable baseline for a first-generation device from a company that has never manufactured hardware, sold through carriers, handled warranty claims, or managed a consumer-scale supply chain. Jony Ive’s device carries the same risk: a company whose experience is in broad-language models trying to become a consumer electronics manufacturer, a transition that requires competencies in industrial design, supply chain management, retailing and after-sales service, which OpenAI doesn’t have and can’t achieve by hiring a designer, no matter how talented.

The 2028 timeline gives OpenAI two years to finalize specifications, secure component supplies, build manufacturing capacity, develop an agent-first software platform, negotiate carrier partnerships, establish retail distribution and convince hundreds of millions of consumers to ditch their iPhones and Galaxy phones for a device made by a company that has never shipped hardware. Humane AI Pin lasted longer than that, shipping a device that lasted nine months before becoming permanently disabled. Ambition is extraordinary. The supply chain is reliable. The concept addresses a real architectural limitation of existing smartphones, which were designed around apps in 2007 and haven’t changed fundamentally since then. But the distance between a reliable supply chain report and a shipping product that displaces an iPhone is the distance between a thesis and a business, and every company in the AI ​​device graveyard has a thesis.