SpaceX has reportedly shown a prototype of a handset-like AI device to investors ahead of its record-breaking initial public offering, according to a recent report from The Wall Street Journal. The device, which is described as slimmer than an iPhone, would run on a proprietary operating system, integrate technology from xAI—the artificial intelligence company founded by Elon Musk—and use a Qualcomm Snapdragon chipset. However, Musk himself has flatly denied the report, calling it “utterly false” on his social media platform X.
The Journal’s sources indicated that the project is still in its early stages, with the design subject to change and no guarantee that the device will ever reach mass production. Despite this caveat, news of the potential hardware caused Qualcomm’s shares to rise by approximately three percent on the day of the report. SpaceX has not issued any official public statement about the device, leaving the market to speculate based on the Journal’s account.
The xAI merger and its implications
In a move that significantly reshaped Musk’s corporate web, SpaceX absorbed xAI in February 2025 in a merger valued at roughly $1.25 trillion. This acquisition gave the rocket company direct access to the advanced AI models and computing infrastructure that Musk’s AI lab had been developing. A proprietary device running xAI’s technology would allow SpaceX to operate completely outside the Android and iOS ecosystems, bypassing the platform fees and software restrictions that come with building on someone else’s operating system.
This integration aligns with Musk’s long-standing criticism of the duopoly held by Apple and Google over mobile operating systems. By creating its own hardware and software stack, SpaceX could retain full control over user data, app distribution, and revenue streams—something that few companies have ever achieved. The Device would also be deeply integrated with xAI’s Grok chatbot and other AI agents, potentially offering a voice-first interface that reduces reliance on traditional app stores.
Starlink and the wireless ambition
SpaceX’s interest in the consumer device market does not exist in a vacuum. The company has been actively building out its Starlink satellite internet constellation, which now has over 5,000 operational satellites. Recently, SpaceX told investors it plans to sell Starlink phone service directly to US consumers, challenging established carriers like Verizon, AT&T, and T-Mobile. The company also acquired valuable wireless spectrum from EchoStar for $17 billion, further solidifying its ability to offer mobile connectivity independent of terrestrial networks.
An AI device designed to work seamlessly with Starlink would give SpaceX a vertically integrated product—hardware, software, and connectivity all under one roof. Users could make calls, send messages, and access AI services anywhere on the planet where Starlink coverage exists, without needing a traditional carrier plan. This would be a powerful selling point for remote areas, travelers, and businesses operating in underserved regions.
The competitive landscape
If the report is accurate, SpaceX would be entering a race that already has a well-funded frontrunner. OpenAI, the company behind ChatGPT, has been quietly developing its own AI-focused smartphone. In early 2025, OpenAI hired Paul Meade, the former Apple vice president who oversaw the Vision Pro hardware engineering team. Meade joins a team that includes Jony Ive, Apple’s former design chief, who has been working with OpenAI on a device that Sam Altman has described as “more peaceful” than an iPhone. OpenAI is reportedly targeting mass production in 2028, using chips from both Qualcomm and MediaTek.
Other major technology companies have also expressed interest in dedicated AI hardware. Meta has invested heavily in augmented reality glasses, while Google continues to evolve its Pixel line with on-device AI capabilities. However, the market for standalone AI devices remains small and littered with failures.
The graveyard of AI hardware
The recent history of AI hardware is stark. Humane’s AI Pin—a wearable device that projected information onto the user’s palm and used AI for voice commands—was permanently bricked in February 2025 after selling fewer than 10,000 units. The company was acquired by HP for $116 million, a fraction of its earlier valuation. The Rabbit R1, a compact device that used a large language model to perform tasks by controlling apps on behalf of the user, attracted 100,000 pre-orders but retained only around 5,000 active users after five months. Both devices failed because they required consumers to carry a second gadget that did less than the smartphone already in their pocket could do.
Other attempts have met similar fates. The Essential Phone, founded by Android co-creator Andy Rubin, struggled to gain traction despite a unique magnetic accessory system. The Light Phone, designed to minimize screen time, remains a niche product. In each case, the challenge of convincing users to adopt a new form factor, learn a new interface, and trust a new ecosystem proved insurmountable.
SpaceX, however, has distinct advantages. Through Tesla, the company has deep experience in manufacturing consumer electronics at scale, including battery packs, infotainment systems, and even the Cybertruck’s steel body. It also has access to the chips needed for on-device compute, via its relationships with Qualcomm and other suppliers. Furthermore, SpaceX’s loyal fan base—which includes millions of Starlink users and Tesla owners—could provide an initial market that venture-backed startups like Humane and Rabbit never had.
Yet Musk’s outright denial creates an unusual tension. Either the Wall Street Journal’s sources are mistaken or motivated by something other than fact, or SpaceX is now walking back a project that it pitched to investors just weeks ago. Neither scenario is particularly reassuring for those trying to judge the device’s future. The denial also complicates any potential partnership discussions with hardware suppliers or software developers, who might now be hesitant to invest resources in a project that the CEO has publicly disavowed.
Historical context: Musk and hardware ambitions
Elon Musk has a long history of ambitious hardware projects that began as rumors and later became reality. The Tesla Model 3 was greeted with widespread skepticism before it became the best-selling electric car in the world. The Starlink satellite internet service was initially dismissed as a fantasy, yet it now serves hundreds of thousands of users across the globe. Even the Humane AI Pin’s failure hasn’t stopped Musk from pushing forward with brain-implant company Neuralink, which recently implanted its first device in a human patient.
However, Musk has also been known to overpromise and underdeliver. The Tesla Semi is still not in volume production years after its reveal. The Cybertruck has faced multiple delays and design changes. And Musk’s notorious tweet that Tesla would achieve Level 5 autonomy by 2020 remains unfulfilled. So when Musk says a report is false, it is not necessarily a sign that the project is dead—it could be a strategic move to manage expectations or to mislead competitors.
Ultimately, the truth may not be known until SpaceX either confirms or denies the device in an official capacity. Given the company’s history of operating in stealth mode—even announcing the Starlink Direct-to-Cell service only after testing had begun—it is possible that a consumer AI device is indeed in the works, but that Musk wanted to stop the Journal from revealing it before SpaceX was ready. Conversely, the report could simply be a misinterpretation of a research project that was never intended for mass production.
Either way, the intersection of SpaceX’s satellite network, xAI’s artificial intelligence, and Musk’s relentless ambition ensures that this story will continue to generate headlines until the company decides to break its silence. In the meantime, investors and consumers alike are left to parse the conflicting signals and to judge whether the next big thing in AI hardware will come from a rocket company—or from the graveyard of failed gadgets.