OpenAI and Apple Patent Wearables That Read Silent Words

OpenAI and Apple each have a US patent application for a wearable that turns words you mouth, without making a sound, into text or speech. The US Patent and Trademark Office published both on 19 November, with back-to-back numbers: US 2026/0369273 for OpenAI and US 2026/0369274 for Apple.

OpenAI’s filing is titled “In-Ear Silent-Speech Detection System”. It describes an earbud that senses how the ear canal moves when the mouth forms words, then generates “a representation of the one or more words”, according to its abstract.

Apple’s filing, “Systems and Methods for Implementing Silent Speech”, describes “a wearable system for facilitating silent conversations” worn on the head. Both describe two people talking silently through the devices. A patent application is not a product, and neither company has announced one.

Business Pill · Distribution advantage explained

An evergreen Business Engineer explainer. It teaches the general idea only and does not discuss these filings.

The key insight: OpenAI and Apple each have a published US patent application for a wearable that turns silently mouthed words into text or speech, and both describe two people conversing without sound.

What OpenAI Filed

FIG. 1 of OpenAI patent application US 2026/0369273 A1, an earbud beside an ear FIG. 1 of OpenAI’s application. Source: USPTO, US 2026/0369273 A1.

The applicant is OpenAI OpCo, LLC, with two inventors, Douglas Stuart Brodie and Daniel Boothe. OpenAI filed it on 14 May 2026, a year after a provisional application dated 14 May 2025, the front page shows.

The sensing sits in the part of the earbud that goes into the ear canal. Its drawings list motion, deformation, vibration and pressure sensing. A machine-learning step then classifies phonemes or words, and the output can be shown on a display, read out by text-to-speech or run as a command.

OpenAI's application was published about 18 months after its first provisional filing and Apple's about 26 mon
OpenAI’s application was published about 18 months after its first provisional filing and Apple’s about 26 months after its earliest, both on 19 November 2026, by our arithmetic on the USPTO front pages.

Two People, No Sound

The filing spells out a silent conversation. “A representation of words that were silently spoken by a first user may be sent to a second user’s ear-worn sensing device to be output as a spoken output,” it says. The second user’s silent reply is processed and sent back the same way.

The words can also go to a phone, a computer or a head-mounted display, the application says.

What Apple Filed

Apple Inc. is the applicant, with four inventors in Israel and Massachusetts. The application, filed 8 July 2026, continues an international filing from September 2025, and its earliest provisional is dated 9 September 2024.

Its sensor reads “communication-related neuromuscular activity”, the abstract says. The drawings describe a system that “detects facial skin micromovements” and interprets them “from light reflections”, with a non-visible light source. The post that flagged the filings described it as laser light on the cheek.

One drawing shows “two individuals conducting silent communications using wearable systems”. In it, one person thinks “Coffee later?” and the other receives it and answers “Sure. Meet me at the Corner Café.”

Why It Matters

Both filings describe talking to a device or another person without making a sound. Silent speech is already a start-up category; Subvocal raised $1.3M for $499 silent-speech glasses, as we reported on 30 October.

The two applications were published on the same day. OpenAI’s took about 18 months from its first provisional and Apple’s about 26, by our arithmetic on the dates on each front page.

The Structural Read

As we read it, both filings treat silent speech as a way to talk to an assistant in public without being heard.

As we read it, the earbud in OpenAI’s filing works as an input device, not only a speaker.

OpenAI patent application US 2026/0369273 A1

“A representation of words that were silently spoken by a first user may be sent to a second user’s ear-worn sensing device to be output as a spoken output.”

Three Implications

FOR OPENAI an in-ear device that senses ear-canal motion when the mouth forms words, per its application.

FOR APPLE a head-worn system reading facial skin micromovements from light reflections, per its application.

FOR READERS neither application is examined or granted, and no product has been announced.

What Is Not Established

Patent applications describe what a company wants to protect, not what it will ship. Neither has been examined or granted, and neither OpenAI nor Apple has announced a silent-speech product. The filings give no accuracy figures that we read.

This Week In The AI Supercycle

A World Without an Order, and an AI Central Bank Without a Successor

This week’s issue: the week AI memory stayed largely sold out while the people holding the shares paid: why the AI Supercycle is not a dot-com replay, read through geopolitics, macro and technology, with the gap between building and financing the widest we’ve recorded.

Read the full issue →

Business Engineer Framework

The AI Supercycle

The Business Engineer map of where AI capability turns into business leverage.

Read the Map of AI →

The Bottom Line

OpenAI’s earbud reads the ear canal and Apple’s head-worn device reads facial skin with light. Both applications, published on 19 November, describe two people holding a conversation without making a sound.

97,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.

A note on sourcing. Titles, applicants, inventors, dates, abstracts and quoted passages come from the two USPTO publications, read on 21 November 2026. The 18 and 26 months are our arithmetic on the dates printed on each front page. The applications have not been examined, and the companies have not commented on them that we have seen.

Sources: USPTO, US 2026/0369273 A1, ‘In-Ear Silent-Speech Detection System’, applicant OpenAI OpCo, LLC, published 19 Nov 2026 · USPTO, US 2026/0369274 A1, ‘Systems and Methods for Implementing Silent Speech’, applicant Apple Inc., published 19 Nov 2026 · IntCyberDigest on X, 21 Nov 2026 (flagged both filings) · Patentlyze weekly patent roundup, 21 Nov 2026

Scroll to Top
FourWeekMBA