OpenAI’s AI-first phone could enter mass production by 2028
OpenAI is developing a smartphone built around AI agents instead of apps, according to Ming-Chi Kuo. The analyst says Qualcomm and MediaTek are jointly designing the custom processor, while Luxshare Precision Industry would co-design and exclusively manufacture the device, with mass production targeted for 2028.
OpenAI is reportedly building a smartphone that is designed around AI agents rather than traditional apps, and the supply chain may already be lining up behind it. According to analyst Ming-Chi Kuo, the device is being developed with Qualcomm and MediaTek working on a custom processor, while Luxshare Precision Industry would co-design and exclusively manufacture the phone, with mass production targeted for 2028.
If that timeline holds, this would be one of the clearest signs yet that agent-based computing is moving from software demos into dedicated consumer hardware. Instead of opening separate apps for messaging, search, scheduling, or shopping, the phone would center on an AI system that can carry out tasks on the user’s behalf.
⚡ New to this?
This is about a phone that would be built around AI agents, which are software systems that can plan and carry out tasks for a user instead of just answering single prompts. A custom processor is a chip designed for a specific device, and in this case it would likely be tuned for constant AI work rather than ordinary app use.
Non-experts should care because phones are where most people already interact with digital services, and a shift away from apps could change how software is designed and used. If the report is right, it also shows that AI is moving from chatbots and desktop tools into dedicated consumer hardware, with new implications for privacy, cloud reliance, and device management.
🦞 OpenClaw angle
For AI automation builders, this is a signal that agent-first interfaces may move from software experiments to dedicated hardware. For IT and security teams, it raises new questions about device context capture, cloud dependence, and how to manage a phone that is designed around continuous AI inference.
Kuo’s report matters because he has a long track record of tracking Apple’s hardware plans and broader electronics supply chains. He often surfaces details from component makers and manufacturing partners before products are announced, so his notes are closely watched even when they are not official company statements.
The idea of an app-light or app-free phone is not new, but it has usually stalled because current AI systems are still too slow, too unreliable, or too dependent on cloud services. A dedicated device changes the equation by treating the assistant as the core interface, not just an add-on feature tucked inside a phone operating system.
That kind of product would need a custom chip designed to handle frequent AI inference, the process of running models to interpret requests and generate responses in real time. It would also need tight integration between local hardware and cloud services, since many AI agents still rely on remote servers for heavier reasoning, retrieval, or multimodal tasks.
The reported involvement of Qualcomm and MediaTek is notable for another reason. Both companies already supply chips to the broader smartphone market, and their participation suggests OpenAI may be trying to build a device with the kind of performance, power management, and radio support expected of a mainstream handset rather than a niche demo gadget.
Luxshare’s reported role also points to a serious manufacturing plan. The company is a major electronics assembler with experience producing consumer devices at scale, which is the sort of partner a new hardware project would need if it is aiming beyond a limited prototype run.
A 2028 mass-production target is still far enough away that the project could change substantially before launch. Hardware programs often shift on industrial design, component sourcing, operating system choices, and the economics of getting a new phone into a crowded market.
Even so, the outline described by Kuo suggests OpenAI is not just exploring a companion device or smart accessory. It would be trying to build a phone where the AI agent is the primary way users interact with the device, with the processor, manufacturing, and software model all shaped around that goal.
