Kuo: OpenAI builds smartphone with MediaTek & Luxshare Apfelpatient

Apple analyst Ming-Chi Kuo has revealed explosive new details about OpenAI's smartphone project. According to him, the company is working with MediaTek, Qualcomm, and Luxshare on an AI-powered phone that is expected to fundamentally revolutionize the smartphone industry. Mass production is slated to begin in 2028 – and Apple should take this development very seriously.

That OpenAI is working on its own AI device has been known for months – not least due to its acquisition of Jony Ive's hardware startup. However, what exactly is being built remained speculation for a long time. Some reports spoke of a display-less PIN device, others of a similar device, and still others of a smartphone. Ming-Chi Kuo, one of the most reliable analysts for Asian supply chain data, has now provided the first concrete evidence: It is indeed intended to be a fully-fledged smartphone – thus directly challenging Apple's core business.

Kuo designated The project is dubbed the “AI Agent Phone” and described as a fundamentally different kind of smartphone. Instead of a collection of apps, the device is intended to offer a seamless AI agent experience – taking over tasks for the user instead of providing apps to click on. Three partners are behind the project: MediaTek and Qualcomm are developing the processors, and Luxshare is the exclusive system and manufacturing partner. Mass production is planned for 2028.

The concept: tasks instead of apps

At its core, Kuo's vision of an AI agent phone differs radically from existing smartphones. Today, users open apps to accomplish tasks – a weather app for forecasts, a mail app for emails, a maps app for navigation. The app is the central user interface paradigm. The AI agent phone aims to turn this approach on its head.

Kuo puts it this way: users don't want to use "a collection of apps"—they want to accomplish tasks and satisfy needs. This is precisely where the AI agent phone comes in. Instead of opening apps, the user formulates a request, and the AI agent automatically handles the necessary steps. Adding an appointment to the calendar, reserving a restaurant, starting navigation, replying to a message—all via a central, AI-powered interface.

Why OpenAI is building its own smartphone

Kuo cites three key reasons why a smartphone is strategically indispensable for OpenAI. First, only those who fully control the operating system and hardware can deliver a seamless AI agent experience. Those who rely on iOS or Android must adhere to the platform's rules – this significantly limits the capabilities of an AI agent.

Secondly, the smartphone is the only device that practically accompanies the user constantly. Location, activities, communication, health data – everything runs through the device. According to Kuo, this real-time context is precisely the most important input for AI agents. Without this context, the agent cannot work effectively.

Thirdly: Smartphones will remain the most important device category for the foreseeable future. Despite all the discussions about smart glasses, wearables, or other form factors, no other device achieves the reach and depth of use of a smartphone. Therefore, anyone who wants to make an AI agent suitable for mass adoption must focus on the smartphone.

The hybrid model: Cloud meets on-device

Technically, Kuo envisions a close integration between device AI and cloud AI. The processor must continuously understand the user context, execute basic AI models locally, and operate in an energy- and memory-efficient manner. More complex, computationally intensive tasks are handled in the cloud.

This is a proven model that other AI hardware manufacturers also follow. Local processing is fast and saves battery power, while the cloud handles the heavy lifting. However, this concept would take on a special dimension with OpenAI's smartphone – after all, the company owns some of the world's most powerful AI models, which are available in the cloud.

The strengths of OpenAI

According to Kuo, OpenAI brings three key advantages to the table. First, a strong consumer brand. ChatGPT is one of the fastest-growing consumer products in tech history. Second, years of collected user data enable the training of highly accurate models. Third, leading AI models that are continuously being developed – as recently demonstrated with GPT-5.5.

Hardware development is considered a manageable risk for OpenAI because the smartphone hardware industry is already highly standardized. With MediaTek and Qualcomm, OpenAI has two of the most experienced chip developers on board. With Luxshare, the company gains a manufacturing partner that is among the leading contract manufacturers in Asia and already produces high-quality consumer electronics.

The business model: subscription plus hardware

Kuo's point about the business model is particularly interesting. OpenAI apparently plans to bundle ChatGPT subscriptions with the hardware. Those who buy the smartphone will therefore potentially receive an integrated Plus or Pro account. Additionally, a developer ecosystem is to be created around AI agent functions – similar to how Apple built its App Store, but for AI agents instead of apps.

This is an ambitious strategy. OpenAI would not only challenge Apple and Samsung, but also establish a completely new business model: hardware sales plus software subscriptions plus platform fees from developers.

What Luxshare gains from the deal

For Luxshare, the project is of paramount strategic importance. While the Chinese company is one of the largest contract manufacturers in the industry, it has been striving for years to overtake market leader Foxconn (Hon Hai). In OpenAI's smartphone project, Luxshare is the exclusive co-design and manufacturing partner – essentially what Foxconn is for other manufacturers.

Kuo describes this as "particularly significant" for Luxshare. Being positioned early in the supply chain of a new smartphone type allows companies to become market leaders in that new generation. Should the AI-powered agent phone prevail as the next generation of smartphones, Luxshare would be one of the biggest beneficiaries.

What this means for the smartphone industry

If Kuo's information is correct, a new chapter in smartphone history will begin in 2028. For the first time since the iPhone in 2007, a player would emerge who fundamentally rethinks the user interface concept – not just improving details. Instead of app collections, an AI agent; instead of touch-first interfaces, a dialogue-based system.

Whether this will work remains to be seen. Smartphone manufacturers have repeatedly tried to establish alternative operating paradigms over the past 15 years – for example, with voice assistants as the central interface or with gesture-based interfaces. So far, none of these have prevailed against the app model. But OpenAI has something that previous challengers lacked: an AI model that can actually handle complex tasks.

For established smartphone manufacturers like Apple, Samsung, and Google, this means that the next two years will be dedicated to preparation. Those who don't have a convincing answer to the AI agent concept by 2028 could fall behind. However, there are still many steps between Kuo's supply chain analysis and a finished product, and a lot can go wrong along the way. Until OpenAI actually brings a competitive smartphone to market, it remains an extremely ambitious project. (Image: Shutterstock / Henry Franklin)