OpenAI Bets $7B on Building Its Own AI Phone

OpenAI has made its boldest move yet into consumer hardware, acquiring legendary designer Jony Ive's startup IO Product at a valuation of nearly $7 billion. The deal brings roughly 50 former Apple engineers — many of whom helped build the original iPhone and iPod — into OpenAI's orbit, with a singular mission: creating a dedicated hardware device purpose-built for ChatGPT and large language models.

Court filings from a US district court have revealed text messages and emails from key team members, including Tang Tan, expressing intense excitement about AI hardware's potential. Their tone, according to observers, mirrors the infectious optimism of startup founders who believe they are standing at the edge of a transformational moment in computing.

Key Takeaways

• OpenAI acquired Jony Ive's IO Product at a valuation of nearly $7 billion
• The team includes roughly 50 former Apple engineers who worked on the original iPhone and iPod
• Court filings reveal the team's vision for a dedicated ChatGPT hardware device
• The engineers explicitly avoided calling the project an 'iPhone killer'
• OpenAI previously connected another hardware startup with IO Product before the acquisition
• The core thesis: ChatGPT needs its own 'beautiful box,' just as iOS needed the iPhone

Former iPhone Engineers Chase the Next Computing Paradigm

The engineers behind this project are not typical startup founders. Over the past 2 years, key members of Apple's hardware team — people who witnessed and directly contributed to the creation of the iPod and iPhone — left the company one by one. They followed Jony Ive, Apple's former chief design officer and the man Steve Jobs once called his 'spiritual partner' in design.

At IO Product, this team began exploring what a post-smartphone computing device might look like. Their emails, now part of court records, reveal a team that is deeply convinced AI represents the most significant hardware innovation opportunity since the smartphone revolution began in 2007.

Notably, these veterans are careful about framing. Despite their pedigree, they deliberately avoid positioning their work as an 'iPhone killer.' Having lived through the iPhone's rise from the inside, they understand better than anyone that the next paradigm shift won't come from simply replacing the smartphone — it will come from reimagining what a personal computing device can be when AI is at its core.

Why ChatGPT 'Needs Its Own Box'

The philosophical foundation of the project traces back to a famous Steve Jobs insight: the iPhone was never the product — it was merely a beautiful box for iOS. The hardware existed to contain, constrain, and showcase the software experience.

OpenAI's leadership appears to have embraced this logic wholeheartedly. ChatGPT, they believe, needs its own dedicated 'box' — a purpose-built device that can:

• Constrain the AI experience into a focused, coherent interaction model
• Guide the development trajectory of LLM technology through hardware-software co-design
• Differentiate OpenAI's offering in a market where large language models are increasingly commoditized
• Establish a direct consumer relationship that bypasses Apple, Google, and other platform gatekeepers

This is perhaps the most strategically important point. Today, ChatGPT lives inside other companies' ecosystems — as an app on iPhones, as a browser tab on Windows PCs. OpenAI captures none of the platform economics that have made Apple and Google the most valuable companies in the world.

The Commoditization Problem: Can Hardware Save AI?

The source material raises a provocative question that sits at the heart of OpenAI's hardware bet: right now, there appears to be no meaningful differentiation between major LLMs. Unlike Google Search, which built an unassailable moat through data and algorithmic advantages, or Windows, which locked in users through software compatibility — no AI model has yet achieved true platform dominance.

GPT-4o, Claude 3.5 Sonnet, Gemini 1.5 Pro, and Llama 3.1 are all remarkably capable, and the gap between them continues to narrow with each release cycle. If the models themselves cannot be meaningfully differentiated, the value must be captured elsewhere — in distribution, in user experience, or in hardware.

Apple proved this thesis decades ago. The iPod was not the first MP3 player. The iPhone was not the first smartphone. But Apple's genius was in creating hardware that made the software experience feel magical, proprietary, and irreplaceable. OpenAI appears to be making the same bet: that a purpose-built AI device can create the kind of user lock-in that the models alone cannot.

This is a high-stakes gamble. The graveyard of 'next computing platform' devices is vast — from Google Glass to the Essential Phone to Humane's AI Pin, which launched to devastating reviews in early 2024. Even Meta's multi-billion dollar bet on VR headsets has struggled to achieve mainstream adoption.

How the Acquisition Came Together

The path to OpenAI's acquisition of IO Product reveals interesting dynamics in the AI hardware ecosystem. According to court filings, a separate hardware startup had initially approached OpenAI seeking investment. Rather than funding them directly, OpenAI introduced the team to IO Product.

During the subsequent email exchanges, the former Apple engineers at IO Product shared their vision for AI hardware with remarkable candor. These communications — now public through legal proceedings — paint a picture of a team that sees AI hardware not as a niche accessory but as the next dominant computing platform.

The $6.9 billion valuation for a 50-person team is extraordinary by any measure. It works out to roughly $138 million per employee — a figure that reflects not just the team's talent but OpenAI's strategic urgency. For comparison, when Google acquired Nest in 2014, it paid $3.2 billion for a company with over 200 employees and actual products on the market.

OpenAI's willingness to pay such a premium suggests the company views dedicated hardware as existentially important to its long-term strategy, not merely a 'nice to have' product extension.

What This Means for the Industry

OpenAI's hardware push has significant implications across the tech landscape:

• For Apple: A team of its own alumni, led by its most iconic designer, is now building a competing device. This represents both a talent drain and a potential strategic threat to the iPhone's dominance as the default AI interface.
• For Google and Samsung: Android device makers must now consider whether OpenAI's hardware could fragment the AI assistant market further, beyond the existing Gemini vs. ChatGPT competition.
• For developers: A new hardware platform could mean a new app ecosystem, new APIs, and new design paradigms — potentially creating early-mover advantages for those who build for it.
• For consumers: The promise is a device where AI is not an add-on feature but the fundamental organizing principle of the entire user experience.
• For AI startups: OpenAI's vertical integration strategy — from model training to consumer hardware — raises the bar for competition dramatically.

The move also signals a broader industry trend. AI companies are increasingly looking beyond software to capture value. Whether through hardware (OpenAI), through enterprise platforms (Microsoft Copilot), or through cloud infrastructure (Google, Amazon), the race is on to build durable competitive advantages around AI capabilities that are themselves becoming commoditized.

Looking Ahead: Timeline and Challenges

No official timeline has been announced for an OpenAI hardware device. However, given that IO Product was already operational with a 50-person team before the acquisition, it is reasonable to expect a prototype or announcement within 12 to 18 months.

The challenges ahead are formidable. Hardware is fundamentally different from software — it requires supply chain management, manufacturing partnerships, regulatory compliance, and retail distribution. These are capabilities OpenAI has never needed before.

Moreover, the team must answer the fundamental design question: what form factor does an AI-native device take? Is it a phone? A wearable? Something entirely new? The failures of Humane's AI Pin ($699, widely panned) and Rabbit's R1 (limited functionality) suggest that simply removing the screen or reimagining the smartphone is not enough.

What the IO Product team has that those startups did not, however, is decades of experience building devices that hundreds of millions of people actually want to use. Combined with OpenAI's AI capabilities and $13 billion in Microsoft backing, this may be the most credible attempt yet at creating a post-smartphone computing device.

The question is no longer whether AI will get its own hardware — it is whether OpenAI can build the device that makes the AI era feel as inevitable as the iPhone made the smartphone era feel in 2007.