SpaceX Reportedly Sought to Acquire AI Coding Startup Cognition — CEO Says No
Bloomberg reported on August 19, 2026, that SpaceX attempted to acquire Cognition, an AI coding startup. Cognition CEO Scott Wu pushed back on the report the same day, posting on X that the story was inaccurate, that Cognition is not for sale, and that the two companies have not been in talks. TechCrunch
Wu's denial came shortly after the Bloomberg report went live. Bloomberg also reported that while acquisition talks are no longer active, the two companies are still discussing a potential working relationship — possibly with Cognition using SpaceX's computing capacity. TechCrunch
Wu did not address that second claim. His public denial was limited to the acquisition report and the existence of talks. He said nothing about the possibility of a working relationship involving SpaceX's computing infrastructure.
The report surfaced just days after SpaceX completed a $60 billion stock acquisition of Cursor, another AI coding startup, with that deal closing the prior week. The Cursor acquisition came roughly three months after Cognition raised $1 billion in a late-May 2026 funding round at a $25 billion post-money valuation. Bloomberg separately reported that Cognition is now in early talks for a new funding round at a $40 billion valuation. Bloomberg
The broader context here involves two converging dynamics. First, AI coding capabilities are consolidating within companies that already control significant computing resources — the physical infrastructure needed to train and run large AI models. Second, valuations for companies in this category are climbing fast. SpaceX's reported interest in Cognition follows its $60 billion Cursor acquisition, suggesting a deliberate push to integrate AI coding tools. Cognition's trajectory from a $25 billion valuation in May to early talks at $40 billion less than three months later reflects intense investor demand for AI coding startups.
In my view, the unaddressed computing-capacity claim is the more substantive element of this story. An acquisition denial is a straightforward factual dispute. But the prospect of Cognition gaining access to SpaceX's computing resources points to a structural question facing AI startups: access to infrastructure at scale. As AI models grow more complex, the computing power needed to run them becomes a determining factor in which companies stay competitive. Whether through acquisition, partnership, or infrastructure-sharing agreements, the alignment between AI coding tool developers and compute-rich companies is worth close attention.
The $60 billion valuation for the Cursor acquisition and the reported $40 billion target for Cognition's new round suggest the market is pricing in substantial future value for AI coding startups. The speed of Cognition's valuation increase — from $25 billion to a potential $40 billion in under three months — reflects the intensity of investor interest in this category.
The conflict between Bloomberg's sourcing and Wu's denial leaves the factual record partially resolved. The acquisition attempt is disputed by the CEO of the target company. The reported discussions about computing capacity remain unaddressed by Cognition. The early talks for a $40 billion funding round, if they proceed, would provide additional capital independent of any SpaceX involvement. The core fact for readers to track is whether the computing-capacity discussions turn into a formal partnership. TechCrunch
