Musk vs Altman Feud Reignites Over Apple's OpenAI Suit
Elon Musk and Sam Altman traded scam accusations on X after Apple sued OpenAI. Here's the context: dueling IPOs, the model race, and what's really at stake.
The two men who co-founded OpenAI a decade ago spent the weekend calling each other frauds in public. On Friday, July 10, 2026, Apple filed a trade-secret lawsuit against OpenAI and two former Apple employees, alleging they stole company secrets to accelerate OpenAI’s push into consumer hardware. Within hours, Elon Musk seized on the news to attack Sam Altman, and by the following morning the exchange had escalated into the sharpest round yet of a rivalry that now spans two of the most valuable private-then-public companies in technology.
The spat is easy to dismiss as noise between two billionaires who no longer speak. But it sits on top of a genuine business collision: competing frontier models shipped days apart, back-to-back record IPOs, and a hardware fight that has now reached federal court. The insults are the surface. The substance underneath is a contest over who leads the next phase of the AI industry.
What set it off
The trigger was Apple’s complaint, filed in the Northern District of California, accusing OpenAI of building its device business on misappropriated trade secrets. The suit names two ex-Apple employees among a larger group; OpenAI has reportedly hired more than 400 former Apple staff, many from the teams behind Apple’s M-series chip architecture and its shelved self-driving car program. We covered the filing in detail in our report on Apple’s trade-secret lawsuit against OpenAI.
Musk, who left OpenAI’s board in 2018 and has litigated against the company himself, used the moment to revive a familiar jab. “Scam Altman strikes again,” he wrote on X, recycling a nickname he has deployed against Altman before. Altman did not let it pass. He fired back by pointing at Musk’s own fundraising pitch, noting pointedly that Musk is “the one selling investors on space datacenters” — a reference to the orbital compute ambitions Musk has folded into his AI story.
The two had already been sparring the prior week. When OpenAI released GPT-5.6 Sol just one day after Musk’s Grok 4.5 launch, Altman posted that “there are a lot of benchmarks that suggest 5.6 Sol is the best model in the world right now, but the most reliable way to tell is that Elon is obsessed with me again.” The line captured the dynamic precisely: every product move is now also a personal provocation.
The companies behind the insults
What makes this more than a social-media grudge is how large both sides have become, and how directly they now compete.
Musk’s AI operation no longer sits in a standalone startup. SpaceX acquired xAI in February 2026, absorbing xAI’s data centers, its Grok models and image generator, and the social network X. The combined entity has been billed as SpaceXAI, with the AI business framed as a core part of the parent company’s future. That framing matters financially, because SpaceX then went public in one of the largest offerings ever recorded.
On June 12, 2026, SpaceX completed the biggest IPO in financial history, raising roughly $75 billion at a valuation near $1.77 trillion. The stock priced at $135, opened at $160.95 — a 19.2% first-day premium — and the company’s market value has since pushed past $2 trillion, briefly making it worth more than Musk’s automaker, Tesla. The shares have been volatile since, peaking above $225 intraday before giving back much of the gain, a reminder that public markets are still deciding how to price an aerospace-and-AI conglomerate.
OpenAI, for its part, confidentially filed for its own IPO in June, setting up a race in which both founders are courting public investors at the same time. That parallel is central to the venom: each man is, in effect, pitching the same pool of capital on a competing vision of who will define AI. The IPO wave now sweeping the sector — from OpenAI to Anthropic’s own confidential IPO filing — has turned a research rivalry into a capital-markets one.
The model race is the real battleground
Strip away the personalities and the fight is about product. Over roughly a week in early July, the frontier tier saw an unusually dense cluster of releases. Musk shipped Grok 4.5, positioned as an Opus-class model; OpenAI answered a day later with GPT-5.6 Sol, the flagship of a new family that also includes the cheaper Terra and the fast Luna tiers. We broke down OpenAI’s lineup in our coverage of GPT-5.6 Sol, Terra and Luna and Musk’s release in our writeup on the Grok 4.5 launch.
The timing was not coincidental. Shipping a flagship model within 24 hours of a competitor is a way of denying the rival a clean news cycle, and both companies have grown fluent in the tactic. Altman’s “best model in the world” claim and Musk’s counterpunches are marketing as much as they are ego — benchmark bragging rights translate directly into developer mindshare and enterprise contracts.
The hardware dimension raises the stakes further. OpenAI’s aggressive hiring from Apple, now the subject of litigation, reflects an ambition to build devices, not just models — an ambition that also runs through its ChatGPT Work desktop application and its broader consumer push. Musk, meanwhile, controls both frontier models and the compute and distribution that come with SpaceX’s infrastructure and X’s audience. Each is trying to own more of the stack than the other.
The history that fuels it
The animosity is not new. Musk helped found OpenAI in 2015 as a nonprofit research lab and left its board in 2018 after a disagreement over direction and control. He has since sued the company over its shift to a capped-profit structure and its deepening partnership with Microsoft, arguing the organization abandoned its founding mission. Altman has largely responded by building — and by needling Musk when the opportunity arises.
What has changed is scale. In 2018 the dispute was between a board member and a small lab. In 2026 it is between two companies with a combined valuation in the trillions, competing for the same engineers, the same chips, the same enterprise budgets, and now the same public-market investors. The personal history gives the feud its edge; the business overlap gives it consequence.
What it means
The insults will keep coming, and they are worth watching only as a signal of where the real pressure is. Three things actually matter here.
First, the hardware fight is the one to track. The Apple lawsuit is not a sideshow — it is a direct challenge to OpenAI’s device ambitions at the exact moment the company is trying to go public. If discovery surfaces evidence that OpenAI’s hardware roadmap leans on Apple’s trade secrets, it could complicate the IPO narrative and hand Musk a genuine, not just rhetorical, advantage. A settlement is the likeliest outcome, but the timing is awkward for OpenAI.
Second, the dueling IPOs put a clock on the rivalry. Both founders are now accountable to public shareholders, or soon will be. That tends to sharpen focus on revenue, margins, and defensible moats rather than benchmark tweets. Expect the messaging to stay loud but the underlying strategy to get more disciplined, because quarterly reporting punishes distraction. SpaceX’s post-IPO volatility is an early taste of how unforgiving that scrutiny can be.
Third, the model cadence favors buyers. The compressed release schedule — Grok 4.5 and GPT-5.6 Sol within a day of each other — is a gift to developers and enterprises, who are getting frontier capability at falling prices as each side races to out-ship the other. The competitive spite that makes for bad manners makes for good pricing. For everyone building on these models, the feud is a feature.
Who wins? For now, the clearest winners are the customers caught in the crossfire and the bankers underwriting both IPOs. The clearest risk is that a personal grudge between two people who control an outsized share of the industry’s compute, capital, and talent starts driving decisions that a colder analysis would not. That is the thing to watch: not the next insult, but the first time one of these companies makes a strategic choice that looks more like spite than strategy.
Keep reading
Chisato · · 6 min read OpenAI GPT-5.6-Cyber: What It Is and Who Gets Access
OpenAI launched GPT-5.6-Cyber and split its Daybreak security program into Blue and Red tiers. What the model does, its benchmarks, and who can use it.
Chisato · · 6 min read OpenAI Astra Solves 10 Open Math Problems With Lean Proofs
OpenAI says an internal version of Astra, its next major model, solved ten long-open math problems — each shipped with a machine-checkable Lean proof.
Chisato · · 6 min read Lilian Weng Rejoins OpenAI to Lead Self-Improvement
Thinking Machines co-founder Lilian Weng left the startup citing health, then rejoined OpenAI within days to lead a new recursive self-improvement research team.