OpenAI is reportedly exploring a smartphone that could upend how users interact with devices, replacing traditional mobile apps with AI agents that can execute tasks on command. The device, still in early development, may enter production around 2028, according to industry reports.
The intended device will change how smartphones work by shifting their navigation from an application-centered model to an intent-based one. The user will not need to access individual apps, such as those for instant messaging, booking, or payments; instead, they will give a command to their personal AI assistant, which will then execute the task.
This is meant to address the challenges of interacting digitally on a day-to-day basis.
Rumors point to a blend between device-based and cloud-based approaches to AI. In this scenario, the device-based solution would handle the user’s personal context, habits, and behavior, while the cloud would handle complex, multi-step processes.
The new dual model will be capable of providing quicker responses and intelligent decision-making. It will also demonstrate the transition towards building AI at the core of the product.
The development will reportedly rely on hardware partners specializing in chip design and device manufacturing. Although OpenAI has not officially clarified whether this information is accurate, it implies a comprehensive approach to combining AI, software, and hardware.
It also means that OpenAI will find itself competing with current mobile ecosystems controlled by Apple and Google.
However, the most important difference is the absence of apps as intermediaries between devices and users. The new phones are likely to be built using an AI agent approach, enabling communication with services without applications.
If implemented, such a change will pose a serious challenge for app stores and software development.
Also Read: OpenAI Trades Chat for Speed, Accuracy and Real Work With New GPT-5.5 Rollout
The idea implies that there will be many issues related to third parties' use and processing of information. An agent needs to collect and process personal information, such as communications and location data.
There are also concerns about the quality of the decisions made by artificial intelligence.
As was mentioned above, the development of the agent-based smartphone was not officially acknowledged by OpenAI. However, it reflects the overall industry trend toward the agent approach in computing.