
This previous spring, former OpenAI government Fidji Simo warned staff the corporate was vulnerable to lacking its second as a result of it was “distracted by side quests,” referring to pricey software program tasks like Sora, which OpenAI shut down in April. This week, Bloomberg‘s Mark Gurman shared an replace on the corporate’s nascent {hardware} efforts, writing that OpenAI’s debut first-party machine can be a “humanlike” rechargeable smart speaker. For an unprofitable firm, this feels just like the epitome of what Simo hoped to keep away from.
In 2026, the marketplace for sensible audio system and shows is not rising. “When it comes to the pattern line, it has been slowing for fairly a while, and we’re anticipating declines within the upcoming yr,” says Jitesh Ubrani, the director of worldwide machine trackers at market analysis agency IDC. He declined to share particular numbers associated to the scale of the sensible speaker market, however did say cargo quantity lately has dwindled to “tens of hundreds of thousands models” following 2022 when the tech giants, led by Amazon, shipped greater than 100 million units.
In response to the agency’s knowledge, shipments declined by 16.3 % year-over-year in 2023. The yr after, they went down by one other 11.8 %. The introduction of Alexa+ in 2025 seems to have lessened the bleeding considerably, however the market nonetheless contracted by 6.7 % that yr. By the tip of this yr, IDC estimates it should shrink one other 9.6 % earlier than flatlining subsequent yr.
Within the US, Ubrani attributes the state of issues to the very fact most individuals aren’t upgrading their sensible audio system. “[If you’re a consumer], there’s actually no must refresh your machine as a result of the newest options are cloud-based, and so you’ll be able to proceed on with the identical {hardware},” he says. “From a use case perspective, there’s not an entire lot individuals are doing with these units. They use them for music, podcasts, timers, these types of issues. It is quite simple compute.”
For OpenAI, Ubrani believes making a dent out there will probably be troublesome. “Creating the machine shouldn’t be laborious, however getting it to succeed in any form of mass scale, that is much more troublesome to do as a result of there are established gamers on the market, and the class itself has not developed an entire lot,” he says.
One main difficulty may very well be the value of the machine itself. Bloomberg describes a sensible speaker with a digicam and sensors to assist it perceive its environment, in addition to a chargeable battery and “mechanical parts that may transfer on their very own.” That sounds much more difficult (and costly) than say the standard Echo Dot.
“Sensible audio system for a few years had been a really low-cost machine. We’re speaking about underneath $100, and people had been the units that offered the most effective,” says Ubrani. “We have seen pricing transfer up, however nobody views this as a premium class. Many individuals would balk on the thought of spending $300 for a speaker, and whoever enters the class has to have extra reasonably priced choices accessible.”
In response to Bloomberg, OpenAI “believes the product’s defining function will probably be its persona and talent to attach on a humanlike stage with customers.” The machine will leverage the corporate’s new GPT-Live-1 voice model, which is constructed on a duplex structure, giving it the flexibility to concurrently course of inputs whereas producing an output. “This enables the mannequin to interact in a extra pure back-and-forth, preserve a greater sense of time, and even carry out reside translations,” OpenAI stated of the tech lately.
Amazon made the same set of claims when it announced Alexa+ last year. Because of generative AI, the corporate stated the brand new digital assistant may detect the consumer’s tone and temper and adapt accordingly. It additionally affords contextual consciousness, giving it the flexibility to “bear in mind” earlier components of a dialog. Regardless of these enhancements, Alexa+ does not seem to have moved the needle for Amazon. “Even after the introduction of the smarter AIs like Alexa+ and Gemini, we have not seen an enormous change within the trajectory of the market,” says Ubrani.
It is also unclear if OpenAI may even earn cash off a sensible speaker. Once more, let us take a look at Amazon, the dominant participant within the area. As of 2022, the corporate had didn’t discover a option to generate ongoing income from Alexa, in accordance with a report Business Insider printed that yr. On the time, the corporate’s Echo household of sensible audio system had been among the many best-selling merchandise on its market, however Amazon offered most of these units at price and had didn’t discover a option to monetize interactions with Alexa.
Per Enterprise Insider, a significant downside was that straightforward requests, like a command to play a music observe or a query in regards to the climate, represented the vast majority of what folks had been utilizing Alexa for, and fulfilling these requests made little to no cash for Amazon. In the course of the first quarter of 2022, Amazon’s Worldwide Digital unit, which included Alexa, Echo units and Prime Video, reportedly had an working lack of $3 billion, with the “overwhelming majority” tied to the digital assistant. Issues might need modified for Amazon since then, however aside from asserting that 1 million people had been utilizing Alexa+ as of June 2025, the corporate hasn’t shared many notable milestones since then.
“I do not know if anybody makes cash on the {hardware} or not,” says Ubrani of the opposite main gamers out there. “[Companies] primarily use these units as a automobile to their ecosystem.”
Regardless of these considerations, Ubrani says there’s not less than a technique OpenAI’s plan is sensible. “The business as an entire agrees AI is the best way ahead, however AI cannot simply be locked in a browser. It has to grasp the world round you to be actually useful, and the one method an organization like OpenAI can do that’s if they’ve {hardware},” he says. “It has to start out someplace, and it is a option to begin. It is also a option to enter the {hardware} market with the least quantity of resistance.” He factors out constructing a smartphone or a pair of sensible glasses can be dearer because of the extra difficult provide chains concerned in each classes, in addition to the price of provider certification for cell units.
Nonetheless, Ubrani agrees making {hardware} is troublesome, and there are current examples of corporations failing within the precise market OpenAI is trying to enter. Most famously, Meta ended growth of its Portal smart displays in 2022 in a transfer that noticed it lay off 11,000 people on the similar time. And it is value repeating, OpenAI shouldn’t be worthwhile. Leaked monetary paperwork present the corporate lost $5.09 billion in 2024 and $38.5 billion in 2025, and OpenAI itself has stated it does not anticipate to be worthwhile until at least 2030 due to surging AI costs. A sensible speaker will not be what helps it flip the nook.