OpenAI is designing multiple AI devices, but won't say what they are yet

OpenAI president Greg Brockman announced in a recent interview that the company is actively building a range of hardware products designed to work with its AI models. While he confirmed the initiative exists, he remained deliberately vague about timelines and specifications.

What Brockman revealed#

In an interview with Joanna Stern, Brockman indicated the devices would arrive "soon" but declined to pinpoint a launch date. He also sidestepped questions about whether the products include a smart speaker rumored for 2027 or other form factors like the wearable AI devices that have circulated in industry gossip.

When pressed about potential conflicts with Apple—which is currently suing OpenAI—Brockman deflected, saying, "We are focused on our own development and technology," and reassured that the company has "no interest in other companies' trade secrets."

The vision: voice over typing#

Brockman sketched out a broader shift in how people will interact with computers. He expects that voice interfaces will become dominant for most tasks, replacing typing as the primary input method. Acknowledging the practical challenge of shared office spaces, he noted the emergence of soundproof "beak" masks as evidence that the market is already searching for solutions to the ambient noise problem, even if current workarounds feel awkward.

Trust and transparency#

Beyond hardware, Brockman signaled OpenAI's intent to tackle user privacy concerns—particularly around the privacy guarantees of ChatGPT's "temporary chat" mode. The company is developing technical mechanisms to provide verifiable, auditable proof that only the AI system can access private conversations, offering users concrete assurances rather than just promises.