Cognition CEO Scott Wu publicly denied a Wednesday report that Elon Musk's SpaceX attempted to acquire his AI coding startup, writing on X that the story was inaccurate and that the two companies have not been in talks. The denial landed within hours of the initial report, which cited unnamed sources and described SpaceX as pursuing Cognition to accelerate its push into AI coding after falling behind OpenAI, Anthropic, and Google.
Cognition, the maker of the Devin coding agent, raised $1 billion in late May at a $25 billion post-money valuation and is now in early talks for a new round at a $40 billion valuation. Wu's statement that Cognition is not for sale carries weight against that backdrop: the company is one of the last large independent AI coding startups still uncoupled from a frontier model lab.
SpaceX closed its $60 billion acquisition of Cursor last week, a deal that gave the rocket and satellite company a foothold in the fastest-monetizing corner of applied AI. SpaceX also absorbed xAI earlier this year and went public in June, with its market capitalization touching $2.3 trillion at the peak. The pitch to investors leaned heavily on AI ambitions, including a stated plan to build data centers in space.
Key facts
- 01Cognition CEO Scott Wu says the startup 'is not for sale' and denies any acquisition talks with SpaceX.
- 02Cognition raised $1B at a $25B post-money valuation in late May and is now in early talks for a new round at $40B.
- 03SpaceX closed its $60B acquisition of Cursor last week and went public in June at a peak market cap near $2.3T.
- 04Cognition's Devin serves enterprise customers including Mercedes-Benz, Citi, and Goldman Sachs.
- 05Musk told SpaceX employees that in four or five years, AI will be 99% of the company's value.
The xAI business inside SpaceX has lagged. Grok has drawn attention for the wrong reasons — last year's 'MechaHitler' incident and this year's nonconsensual imagery scandals have complicated the enterprise sales motion. Cursor and SpaceX jointly released Grok 4.6 this month, a model tuned for coding and multi-step agentic tasks, and adding Cognition would have layered Devin and an enterprise book that includes Mercedes-Benz, Citi, and Goldman Sachs onto that stack.
Musk told SpaceX employees last week that in about four or five years, AI will be 99% of the value of the company. Hitting that target requires substantially more AI revenue than SpaceX currently books, and coding is the clearest near-term path — Anthropic's growth on the back of Claude Code has demonstrated that clearly enough that every large AI operator is chasing the same customer base.
“four or five years, AI will be 99% of the value”— Elon Musk, SpaceX CEO
According to the report Wu is disputing, active deal talks have ended but the two companies are still discussing a commercial arrangement in which Cognition would use SpaceX's computing capacity. SpaceX currently sells compute to other AI players, including Anthropic, until it needs that capacity for its own workloads. Wu did not address the compute-partnership piece in his denial.
Cognition's cultural profile also matters here. After the company acquired the remaining assets of Windsurf last year — following Google DeepMind's $2.4 billion acqui-hire of Windsurf's CEO and top researchers — Cognition laid off 30 employees and offered buyouts to the remaining 200. Staff who stayed faced 80-plus-hour workweeks and six days in the office. That wartime posture would not have clashed with a Musk-owned parent; Musk has said he works up to 120 hours a week and often sleeps on office and factory floors.
The specifics of what happened between SpaceX and Cognition remain contested. One side of the story is on record from a named CEO; the other side rests on unnamed sources, and SpaceX did not respond to requests for comment. Cognition raising at $40 billion, if that round closes, would price the company well above where a pre-emptive acquisition would clear — a valuation-anchored reason for the board to prefer independence regardless of what conversations occurred.
SpaceX has now bought one coding startup for $60 billion and, per the disputed report, tried to buy another. The pattern is what matters: the company is willing to spend acquisition-sized checks to close the gap in AI coding rather than build the products organically inside xAI. If Cognition stays independent and raises at $40 billion, the pool of standalone coding-agent companies large enough to matter shrinks to a handful, and the price of the next attempted deal goes up. For enterprise buyers weighing Devin against Claude Code and Cursor, the ownership question is no longer academic — it is the roadmap.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
