Jony Ive may be working on a ChatGPT smartphone to compete with the iPhone

Smartphone based on agentic AI 

Unlike current smartphones, OpenAI’s attempt will reportedly bring agentic AI capabilities and an entirely new user interface. Kuo says that the company would replace the usual app-based interface with a task stream based on an agentic AI ecosystem.

The new interface would replace the app grid with a list of tasks and their status. The phone will have on-device AI capabilities, but the more demanding tasks will be sent to the cloud.



Some of the main concerns for the new phone are related to the power consumption and the memory hierarchy management. The key consideration for the development of the processor will be related to the idea that the phone will continuously understand the user’s context.

Why the change of heart?

Since OpenAI announced it was acquiring Jony Ive’s startup, io, last year, we’ve been hearing rumors about different devices, including a small speaker, smart lamp, and glasses. However, the company may want to build a smartphone to have more control over the complete user experience and fulfill its agentic AI ambitions.

Kuo says that the only way for OpenAI to deliver a comprehensive AI agent service is to control both the operating system and hardware of a device. The smartphone form factor is also the only way to capture the user’s state completely and in real time, which is crucial for an AI agent interface.

The analyst also says that smartphones will continue to be the most significant device category for the foreseeable future.

A steep challenge

While ChatGPT is still the most popular AI chatbot on the market, OpenAI is facing growing competition. Anthropic’s Claude and Google’s Gemini are growing in popularity and, in some cases, are performing better than ChatGPT’s models. 

There’s also xAI’s Grok and the Meta AI, which are deeply integrated into the existing products of those companies. Similar to Google’s AI Overviews, those integrations are driving users away from ChatGPT.

Even on the smartphone market, OpenAI will have a very difficult task to win customers over. Apple has been integrating ChatGPT in iOS since the launch of Apple Intelligence in 2024, but the upcoming major Siri update will rely on Google’s language models.

Meanwhile, Samsung is partnering with Perplexity, which is deeply integrated into One UI 8.5 and the Galaxy S26 series phones. Users can choose to add Perplexity to the side button on their phones, while Gemini is handling various AI-powered features across the operating system.

Gemini is also becoming a fundamental part of Android, not only on the Pixel devices. Devices such as the OnePlus 15 and Nothing Phone (4a) Pro feature the base Gemini features and customized AI experience, which are also based on Google’s large language models.

Dead on arrival

Unless AI suddenly has a giant leap in capabilities, OpenAI’s concept won’t be more than an intriguing experiment for AI power users. I don’t think the app-based paradigm is changing anytime soon, and even if it does, I doubt a company without experience in hardware would be the disruptor.