SpaceX, the AI and space exploration giant founded by Elon Musk, has reportedly approached Cognition AI Inc., an AI coding agent leader backed by Silicon Valley venture capital icon Peter Thiel, about a potential acquisition, according to sources familiar with the matter. If finalized, this would mark SpaceX's second major acquisition in recent months following its $60 billion takeover of AI coding startup Cursor, as the company moves to bolster its position in the intensifying artificial intelligence application race. The ultimate goal is to hit a trillion-dollar revenue scale by 2030 or sooner, while realizing the bold narrative that AI operations could account for 99% of SpaceX's value in less than five years.
One source noted that Cognition has not actively engaged on the acquisition proposal. The parties are reportedly still discussing collaboration options, including the possibility of granting Cognition access to SpaceX's AI computing infrastructure. The sources requested anonymity as the information has not been publicly disclosed.
After securing Cursor, SpaceX's overture to Cognition underscores a core motive: capturing the most critical enterprise-grade AI application gateway of the AI era, namely software development workflows and the high-quality feedback data they generate. The strategic significance of acquiring both Cursor and Cognition becomes clearer when viewed against the $28.5 trillion "space-AI empire" blueprint outlined in SpaceX's own prospectus. That projection rivals the U.S. nominal GDP of nearly $32 trillion in Q1 2026, with the enterprise AI application market alone estimated at $22.7 trillion, roughly 70% of total U.S. economic output.
After Musk's $60 Billion Cursor Deal, the Enterprise AI Agent Battle Enters the Compute M&A Era
According to a recent media report, Cognition was valued at $26 billion in a May funding round and has since held preliminary talks with institutional investors seeking a valuation of at least $40 billion. Neither SpaceX nor Cognition responded to requests for comment. Cognition CEO Scott Wu responded to the media reports on social media Wednesday, stating that the company is "not for sale right now, and we have not been in any such negotiations."
The acquisition approach highlights Musk's willingness to invest heavily in the AI compute arms race and the global AI application super-competition. SpaceX recently completed its $60 billion acquisition of Cursor, a fast-growing AI coding startup and a key player in the tech industry's "vibe coding" era. Cognition similarly focuses on vibe coding to help enterprises automate large-scale engineering tasks efficiently.
Musk's companies have been working to build more advanced AI application tools to simplify various AI tasks, including enterprise coding, at scale. Musk's AI business, currently known as SpaceXAI, lags behind AI application rivals like OpenAI and Anthropic in enterprise sales and has undergone a series of layoffs and restructuring. Rapidly establishing a strong growth trajectory for the AI business is a key pillar of Musk's growth strategy for SpaceX.
SpaceX has evolved into a diversified tech giant spanning rockets, satellite communications, AI applications, and AI compute infrastructure, boosted by its historic initial public offering this summer. Founded in 2023, Cognition's flagship product is an AI agent named Devin, designed to automate coding processes for engineers, backed by top venture firms including General Catalyst and Peter Thiel's Founders Fund.
Some sources indicate that SpaceX's interest in Cognition stems not only from its technology but also from its business momentum. Agentic workflows and fully automated programming powered by AI agents have become one of the hottest tracks in AI applications, with leading players like Anthropic, OpenAI, and SpaceX devoting significant portions of their businesses to project-level engineering products.
Cognition AI itself is one of the purest plays on the "Agentic Software" investment theme. Its core product, Devin, is positioned as an autonomous AI software engineer capable of independently planning, writing, testing, and delivering production code within real codebases and development environments. Following the acquisition of Windsurf, the company has built a complete AI software engineering platform combining "Windsurf/Devin agent-coordinated development, Devin for cloud-based asynchronous autonomous execution, and Devin for Terminal for complex engineering tasks." Devin acts as the "cloud-based autonomous AI engineer," Devin for Terminal serves as the "local command-line entry point," and Devin Desktop is the desktop IDE/console for managing collaboration between humans and multiple agents, all serving customers and engineering teams.
For example, Mercedes-Benz has deployed the product across its global R&D and IT systems. In a four-week pilot, Devin analyzed over 200,000 lines of COBOL code, compressing an estimated eight-month modernization effort into eight days. Cognition's website indicates partnerships with major enterprises including Mercedes-Benz Group and GE Aerospace, and these commercial relationships could help SpaceXAI enhance its appeal to potential customers.
Scott Wu has consistently indicated no intention to sell the company, both publicly and in discussions with investors. In a May interview with Bloomberg, he stated that a $1 billion funding round "allows us to remain independent and continue to grow as an independent company, which is very important to us."
Buying the Developer Gateway for $60 Billion, Then Chasing Autonomous Software Engineers: Musk's Full-Stack AI Empire
SpaceX's ambitions in AI applications are indeed vast, but precision is required: the $28.5 trillion disclosed in the prospectus represents the company's overall quantifiable total addressable market (TAM), not an AI revenue target. Of that, the AI application software market accounts for $26.5 trillion, including $2.4 trillion in AI compute infrastructure support, $760 billion in consumer subscriptions, $600 billion in digital advertising, and the most critical $22.7 trillion in enterprise applications.
In other words, Musk is not satisfied with merely selling GPU compute or operating Grok. Instead, he is attempting to integrate "NVIDIA GPUs and data centers, Grok foundation models, Cursor developer workflows, Devin-style autonomous agents, X real-time data, Starlink global connectivity, and Starship and orbital computing," transforming SpaceX into a full-stack AI super-platform spanning infrastructure, models, and applications.
The core motivation behind SpaceX's approach to Cognition after securing Cursor is undoubtedly to seize the most critical enterprise gateway of the AI era: software development workflows and the high-quality feedback data they generate. Cursor brings professional developer distribution channels, enterprise seat revenue, and code interaction data. Cognition offers autonomous agents capable of executing complex engineering tasks, along with enterprise relationships at Mercedes-Benz and GE Aerospace. Integrating these applications with Colossus and Grok could improve utilization of owned compute capacity, reduce third-party model and per-token costs, and create a data flywheel of "user tasks, code execution, result evaluation, and model retraining." It could also directly boost SpaceX's internal R&D efficiency in rockets, satellites, communications, and data center engineering.
Cursor is now a wholly owned subsidiary of SpaceX, but regarding Cognition, this is only an unanswered acquisition approach and potential compute collaboration, not a concluded deal. Cognition's flagship business, Devin, is positioned as an "autonomous AI software engineer": it can independently plan, write, test, and deliver production code after receiving objectives, and can execute cross-codebase migrations, troubleshooting, code reviews, and long-term engineering projects. Cursor's core, meanwhile, is an AI-native editor and multi-model workflow platform embedded in developers' daily work environments, emphasizing human-in-the-loop collaboration, codebase retrieval, completion, editing, debugging, and agentic programming.
In simple terms, Cursor is more like an intelligent cockpit that engineers use continuously, while Devin is more like a digital employee that can independently take on tickets and deliver results. However, as Cursor strengthens its cloud-based agents and Cognition integrates Windsurf, the product boundaries between the two are rapidly converging.
Financial data has already demonstrated that SpaceX's AI ambitions are not purely visionary: SpaceX's Q2 AI-related revenue reached $2.6 billion, up 247% year-over-year, with adjusted EBITDA turning positive for the first time at $1.1 billion. Of the company's $18.4 billion quarterly capital expenditures, approximately $15.8 billion went to AI compute infrastructure. Compute capacity rose from 1GW in Q1 to 1.4GW, with targets of exceeding 2GW by the end of 2026 and approaching 10GW by the end of 2027.
SpaceX management projects at least a $100 billion annualized recurring revenue run rate by December 2026, and has pulled its $1 trillion annual revenue target forward to 2030. Even starting from a $100 billion annualized run rate, this implies roughly 78% compound annual expansion, representing a highly challenging internal plan rather than a low-risk financial guide.
At Morgan Stanley, analysts see the market currently assigning only about $12 per share of implied value to SpaceX's entire AI business. They project Cursor's annualized revenue could reach $33 billion by 2030, supporting a $300 base-case price target and a $600 bull-case scenario. However, the market's reaction to massive capital expenditures and long-term targets after earnings shows that investors will not unconditionally accept Musk's grand narrative that "AI operations will account for 99% of SpaceX's value in five years."
What truly determines the market's revaluation of SpaceX is not the $28.5 trillion TAM, but whether compute capacity can come online steadily, whether Cursor integration can improve gross margins, whether the Grok AI application platform can achieve enterprise-grade competitiveness, whether Cognition-style agents can deliver auditable productivity returns, and whether massive M&A and debt issuance can translate into free cash flow.
Comments