OpenAI is working on a smartphone with MediaTek, Qualcomm, and Luxshare, according to a research note from analyst Ming-Chi Kuo, with mass production targeted for 2028. The device would lean on AI agents instead of traditional apps, and OpenAI would co-design its own smartphone chip with MediaTek and Qualcomm while Luxshare handles manufacturing. Specs and the supplier list are expected to be finalized by year-end 2026 or the first quarter of 2027.
The timing slots in behind a separate hardware bet OpenAI has already telegraphed. Chris Lehane, OpenAI's Chief Global Affairs Officer, said earlier this year that the company is on track to announce its first hardware product in the second half of 2026 — widely reported to be a pair of earbuds. A phone shipping in 2028 would be the second, more ambitious act.
The pitch behind it is consumer scale. ChatGPT is closing in on a billion weekly users, and OpenAI has spent the past year looking for a daily-use surface it controls end to end. A phone that ships with ChatGPT as the operating layer is the most direct path from a chat app to an everyday consumer platform.
Key facts
- 01Ming-Chi Kuo says OpenAI is co-designing a smartphone chip with MediaTek and Qualcomm, with Luxshare as manufacturing partner.
- 02Specs and suppliers are expected to be locked by year-end 2026 or Q1 2027, with mass production targeted for 2028.
- 03Chris Lehane said OpenAI's first hardware product will be announced in the second half of 2026.
- 04ChatGPT is nearing a billion weekly users, the consumer base OpenAI wants to reach with a daily-use device.
- 05Kuo says the phone would mix on-device small models with cloud models, replacing apps with AI agents.
Kuo's note frames the strategic logic bluntly: Apple and Google currently control the app pipeline and the system permissions third parties can access, which caps what any AI assistant on iOS or Android can actually do. By owning the hardware stack, OpenAI would not have to ask Apple for microphone access, background execution, or deep system hooks — it would set those rules itself.
“With ChatGPT closing in on a billion weekly users, OpenAI is aiming to bypass the Apple and Google app pipeline entirely and turn agents into the operating layer.”— Jaeden Schafer
The architecture Kuo describes is a split between small on-device models for low-latency, context-aware tasks and larger cloud models for heavier reasoning. The phone would also continuously parse user context to feed agents, giving OpenAI a richer behavioral signal than any third-party app sitting inside iOS or Android can collect today.
OpenAI is not alone in betting that the app icon grid is on the way out. Nothing CEO Carl Pei said at SXSW that apps will eventually go away, and the broader vibe-coding crowd has been pushing the same thesis: that natural-language agents will swallow the workflows currently fragmented across dozens of installed apps. The disagreement is over timing, not direction.
MediaTek and Qualcomm are an interesting pair to recruit. Qualcomm dominates premium Android silicon and MediaTek has been pushing aggressively into AI-focused mobile chips, and a co-design arrangement gives OpenAI custom silicon without having to stand up an Apple-style chip team from scratch. Luxshare, best known as a major Apple supplier, brings the assembly muscle to actually ship a phone at scale.
The hardware push also fits a pattern across OpenAI's recent moves. The company has been building out infrastructure, distribution, and now devices simultaneously, and rivals are following suit — Thinking Machines has been poaching Meta talent, and Anthropic just locked in up to $40B from Google for compute, both of which AI Chat Daily covered earlier this month. Hardware is the next front.
There are reasons to be skeptical that any of this lands on schedule. Phones are brutal: Amazon's Fire Phone failed, and even well-funded AI hardware attempts in the past two years have shipped late, underperformed, or quietly disappeared. A 2028 mass-production target leaves roughly two years for OpenAI, MediaTek, Qualcomm, and Luxshare to align on a custom chip, an OS-level agent stack, a supply chain, and a distribution model — and Kuo's note is an analyst forecast, not a confirmed product roadmap. OpenAI did not comment.
If OpenAI does ship, the strategic question is not whether the phone outsells the iPhone — it won't — but whether it forces Apple and Google to open up. The moment a credible AI-native phone exists, regulators and developers gain leverage to argue that iOS and Android need to grant assistants deeper access. That, more than unit volume, is the prize OpenAI is actually chasing here.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
