TL;DR
- Analysts uncovered supply chain contracts suggesting OpenAI is building an AI-first smartphone... no app grid, fully agentic from the lock screen.
- OpenAI is rumored to be working on five hardware categories: a phone, earbuds, a screenless companion device, smart home hardware, and custom chips.
- The design direction, led by former Apple designer Jony Ive, is what Sam Altman calls "a cabin by a lake" instead of "Times Square."
- Mass production is reportedly targeted for 2028, but the competition alone will accelerate Apple and Google's timelines.
- For small businesses: the near-term wins aren't about buying a new phone. They're about the features that are about to get pushed to the phones you already own.
Have you ever unlocked your phone and immediately felt a little worse about your day?
That's the problem Sam Altman is describing. He compared today's smartphones to walking through Times Square. Every app, every notification, every badge all shouting at you at the same time. His version of what comes next: a cabin by a lake. Calm. Still useful. Just not hostile.
That's the design philosophy behind OpenAI's rumored AI-first phone... and if it actually ships, it changes more than just the hardware market.
What's Actually Been Reported
This isn't an official announcement. But analysts have uncovered supply chain contracts that are hard to dismiss. CNET's coverage puts OpenAI in talks with Qualcomm and MediaTek for chips, which signals a device designed to run real intelligence on-device, not just pipe everything to the cloud. Mass production is reportedly targeted for 2028.
The Jony Ive angle is real and worth paying attention to. Ive spent decades as Apple's Chief Design Officer and was the creative force behind the original iPhone, the iMac, and the MacBook Air. He's been working with OpenAI through his design firm io. When someone with that track record is involved in what the device looks like and feels like, you take the hardware ambitions seriously.
Times Square vs. the Cabin
The design philosophy isn't just a metaphor. It's a structural rejection of how smartphones are built today.
Right now, a phone is a grid of apps. You unlock, you pick an app, you do the thing, you exit, you pick another. That made perfect sense when the phone was a computer in your pocket. But when the phone can actually understand what you mean, the grid becomes friction.
An AI-first phone doesn't give you a grid. You say what you need, and the phone figures out how to get there. "When am I supposed to meet with the contractor?" doesn't require opening a calendar. "What do I need to do today based on my messages with my wife?" doesn't require opening a chat app and reading through a week of threads.
The calendar doesn't disappear... it just stops being the thing you interact with. The AI reads context. You talk to it. It handles the rest.
Jackson made a good point in the episode: this would solve the problem of needing a calendar but never actually using one. I'd go further. A lot of what we do on our phones isn't really about the apps. It's about getting to information or taking action. If the AI can collapse those steps, the grid of apps becomes less of an interface and more of a back-end layer.
Why "AI-First From Scratch" Is a Real Advantage
Apple has been trying to make Siri smarter for over a decade. The problem isn't talent or budget. It's that Siri has to live inside a system built before anyone had a clear picture of what an AI assistant should be able to do. Every new capability has to negotiate with years of prior decisions and backward compatibility requirements.
OpenAI doesn't have that problem. They're starting from zero... which means no tech debt, no inherited design choices, no compatibility tax from 2007.
There's something else worth noting here. When you bring a genuinely new product to market, you self-select your customers by how you position it. Nobody buys an AI-first phone reluctantly. The early adopters are people who actually want that experience. That gives you a tight feedback loop with the right users from day one, which is a massive advantage for building something good fast.
ChatGPT went from zero to household name between October 2022 and January 2023. If you pull up Google Trends and compare ChatGPT to Gemini and Claude over the past three years, the gap is stark. Even in the valleys, ChatGPT's search volume is roughly double its closest competitors. "Let me ChatGPT it" became a verb the same way "Google it" did. That's the brand equity OpenAI is bringing into hardware.
What Happens to Apps
The obvious question: does Instagram work on this phone?
Probably... but differently. The more likely model is that apps expose tools. Specific actions the AI can call on your behalf, based on what you tell it you want. Not unlike how Siri Actions technically work today on iPhone, except the intent-matching would actually be good enough to feel native.
If you've ever worked with Model Context Protocol (MCP) or seen how AI agents connect to external tools, you've seen the pattern. The AI doesn't need to know every step inside every app. It just needs access to the right actions and enough context to know which one fits the moment.
The reason Siri Actions have such low adoption today isn't because the concept is wrong. It's because the intent-matching is too rigid. You have to say the exact right thing for it to work. An agentic phone built on a modern LLM doesn't have that problem. You describe what you want, and it finds the path.
Five Hardware Bets OpenAI Is Rumored to Be Making
The phone is the headline. But it's reportedly one of five hardware categories OpenAI is moving into:
- The phone — AI-first, agentic interface, no home screen grid
- AI earbuds — with cameras for visual context, not just audio
- A screenless companion device — voice and gesture, described as the "third core device" alongside phone and laptop, and the one most aligned with Jony Ive's design aesthetic
- Smart home products — a smart speaker play, but with a real AI brain behind it
- Custom chips — not consumer hardware, but owning the silicon gives them full-stack control
If all five ship, OpenAI would have a credible claim to the ambient computing layer of your life... the territory Apple, Amazon, and Google have been carving up piece by piece for years, but potentially unified under a single AI platform.
That's a long-shot outcome and a lot still has to go right. But the ambition is coherent, and the people they're hiring and partnering with are not messing around.
Want to know which AI tools are worth your time right now, while the hardware race plays out? Start with this free checklist.
What This Actually Means for Small Businesses
The 2028 mass production timeline means you're not buying one of these next quarter. But the near-term effect isn't really about the phone itself.
Competition moves incumbents. Android pushed iPhone. iPhone pushed Android. An AI-first phone from the company that invented ChatGPT pushes Apple to accelerate whatever they're building with Apple Intelligence and the next generation of Siri. Features that were on the "eventually" roadmap just moved to "this year." That means smarter, more capable AI is coming to phones you already own, faster than it would have without OpenAI entering the hardware race.
For small businesses specifically, here's what an AI-native phone phone eventually unlocks that a stack of separate apps can't:
A customer texts you. The phone reads context from your previous conversations, sees you've spoken before, and drafts a reply that sounds like you. No CRM lookup. No app switching. Just handled.
Someone calls your business number. The AI picks up, gathers the basic info, and either handles it or forwards to you with a summary. You stop missing leads because you were in the middle of something else.
Your calendar builds itself from your conversations. You told a client "let's talk Thursday" in a text three days ago. The phone already knows.
Some of this is already possible with tools you can set up today. But the phone makes it frictionless for people who aren't going to build their own automation stack. That's most small business owners.
What to Watch For
A 2028 mass production target likely means a developer or beta device surfaces in late 2026 or early 2027. Watch for that. When it shows up, pay attention to what app developers are asked to do to make their apps "agentic-ready." That's going to be an early signal of what integrations matter, and where the first real small business use cases land.
Also watch whether this forces Anthropic or Google to announce their own hardware. Right now, a search for "Anthropic phone" on Google Trends is flatlined. But this kind of competitive pressure has a way of forcing hands.
Either way, competition is good for the people who use the product. OpenAI entering hardware makes everything better faster, even for iPhone users who never switch.
If you're trying to figure out where AI actually fits in your business strategy right now... not in two years... this quiz is a quick way to find the real bottleneck.