Who Gets to Define the AI Phone: Smartphone Makers or AI Giants?

A Battle Over the Right to Define AI Phones Is Underway

When reports surfaced that OpenAI was exploring proprietary hardware devices and even considering building an AI-native phone, the entire tech industry took notice. This is not simply a story about an AI company crossing into hardware — it raises a fundamental question: Who gets to define the future of AI phones?

On one side are traditional smartphone giants like Apple, Huawei, and Samsung, gradually embedding AI capabilities into existing smartphone frameworks. On the other are AI-native companies like OpenAI and ByteDance, attempting to reimagine what a phone should be starting from AI first principles. These two radically different approaches are colliding head-on, and the outcome could determine the trajectory of mobile computing for the next decade.

Path One: Smartphone Makers' 'AI Addition'

Apple, Huawei, and Samsung represent an incremental approach — layering AI capabilities onto mature smartphone platforms.

At WWDC 2024, Apple launched Apple Intelligence, integrating large model capabilities into system-level features such as Siri, Mail, and photo editing. Huawei has empowered its HarmonyOS through the Pangu large model, pushing forward in scenarios including smart search, AI object removal, and intelligent summaries. Samsung's Galaxy S series has also fully embraced Galaxy AI, highlighting real-time translation and image enhancement.

The advantages of this path are obvious:

• Massive user base: Billions of smartphone users worldwide can seamlessly receive AI upgrades
• Mature hardware ecosystem: Chips, displays, and camera systems have been refined over many years
• Channel strength and brand trust: Consumer loyalty to phone brands far exceeds their familiarity with AI companies

But the problems are equally apparent: this 'AI addition' approach is fundamentally about incremental improvement within an old paradigm. AI becomes a collection of scattered feature points rather than the core logic reorganizing the phone experience. Siri gets a bit smarter, the photo album gains a few AI features, but the fundamental way users interact with their phones remains unchanged.

Path Two: AI Companies' 'Phone Subtraction'

OpenAI's ambitions clearly extend further. According to multiple reports, OpenAI founder Sam Altman previously invested in the AI hardware project Humane AI Pin and later collaborated with former Apple designer Jony Ive to explore entirely new AI device form factors. Now OpenAI is directly considering entering the phone space, driven by a core logic: rather than stuffing AI into a phone, rebuild the phone around AI.

ByteDance is also making moves. With the rapid iteration of its Doubao large model and the massive user behavior data accumulated from its short-video and content ecosystems, ByteDance has the capability to build a mobile experience with AI at its core. Although ByteDance has not publicly announced plans to build a phone, its deep integration in areas such as AI assistants and intelligent content distribution already signals an intent to redefine mobile interaction through AI.

The core tenets of the AI-native path are:

• Interaction revolution: Shifting from touchscreen taps to natural language dialogue, multimodal interaction, and even proactive intelligence
• App restructuring: No longer needing dozens of standalone apps — AI Agents orchestrate and complete tasks in a unified manner
• Deep personalization: AI understands user intent from the system level, delivering truly individualized experiences

However, the challenges facing this path are equally formidable. The market failure of Humane AI Pin serves as a cautionary tale — consumers are not yet ready to accept a device form factor that completely upends existing habits. Hardware supply chain management, global channel development, and after-sales service systems are all areas of heavy lifting that AI companies are not accustomed to handling.

A Third Possibility: Convergence and Competition

Reality is seldom an either-or proposition. The more likely scenario is a cross-pollination and ongoing competition between the two paths.

On one hand, smartphone makers are accelerating deeper AI integration. Apple's partnership with OpenAI is a prime example — some Apple Intelligence capabilities directly call upon ChatGPT, indicating that traditional manufacturers are not opposed to leveraging AI companies' core competencies. Huawei is also ramping up investment in its proprietary large models, striving to catch up in AI capability depth.

On the other hand, if AI companies do build phones, they will most likely not start from scratch but instead seek partnerships with established hardware manufacturers. The OpenAI and Jony Ive collaboration hints at this possibility — a top-tier AI brain combined with top-tier hardware design, pursuing a path of co-definition.

Also worth watching is transformation at the operating system level. Whether it is Google deeply integrating Gemini into Android or Huawei's AI-driven restructuring of HarmonyOS, the operating system is becoming a key battleground in the AI phone debate. Whoever controls the definition of AI capabilities at the OS layer will, to a large extent, control the definition of the AI phone itself.

The Real Question: What Kind of AI Phone Do Users Actually Need?

Setting aside industry competition, user demand will ultimately determine the winner. The current reality is that most consumers still have a vague perception of what an 'AI phone' actually is. AI removing passersby from photos, AI generating a piece of copy, AI translating a sentence in a foreign language — these features are practical but far from indispensable.

The true killer application for AI phones has yet to emerge. It might be a 'super assistant' that genuinely understands your schedule, emails, and social relationships and proactively handles tasks on your behalf. Or it might be an 'intent computing' device that eliminates the need for manual phone operation entirely.

Regardless of the form it takes, one thing is becoming increasingly clear: the battle over who defines the AI phone is fundamentally a contest over whether AI capabilities or hardware experience takes the lead. Smartphone makers want AI to serve the phone; AI companies want the phone to serve AI.

Outlook: 2025 Will Be a Critical Watershed

In 2025, several key variables will come into play simultaneously: whether OpenAI's hardware plans materialize, the actual market reception of Apple Intelligence, the maturity of Huawei's HarmonyOS AI ecosystem, ByteDance's progress in bringing the Doubao large model to devices, and new breakthroughs by chipmakers like Qualcomm and MediaTek in on-device AI computing power.

This battle over the right of definition will not be settled in a year or two, but the direction will become much clearer in 2025. For consumers, competition is always a good thing — no matter who ultimately defines the AI phone, users will be the biggest winners in this clash of visions.