SpaceX's $1.5 Trillion IPO, AI Agents Go Open-Source, and the Real Numbers Behind Enterprise AI

TECH BRIEF

SpaceX Aims for Record-Breaking IPO

Elon Musk's SpaceX is moving forward with plans for an initial public offering that would raise significantly more than $30 billion, according to sources familiar with the matter. The company is targeting a valuation of approximately $1.5 trillion, positioning it close to Saudi Aramco's market value during its historic 2019 IPO, which raised $29 billion but represented only 1.5% of the company's ownership.reuters+1​

SpaceX's management and advisors are pursuing a listing as early as mid-to-late 2026, though the timeline could shift to 2027 depending on market conditions. If the company sells 5% of its shares at the proposed valuation, it would trigger the largest IPO in history at roughly $40 billion.finance.yahoo+1​

Revenue projections underscore SpaceX's growth trajectory: the company expects approximately $15 billion in 2025, expanding to $22–24 billion in 2026, primarily driven by Starlink's rapidly expanding satellite internet service. The IPO proceeds are earmarked partly for developing space-based data centers and acquiring chips for their operation, a concept Musk discussed recently with Baron Capital.reuters+1​

In the current secondary offering, SpaceX has set a share price of approximately $420, allowing employees to sell roughly $2 billion in shares while SpaceX repurchases some stock. The acceleration toward public markets reflects confidence in Starlink's direct-to-mobile potential and Starship development for lunar and Martian missions.finance.yahoo+1​

AI Giants Unite on Open Agent Standards

OpenAI, Anthropic, and Block announced the formation of the Agentic AI Foundation under the Linux Foundation, marking a significant shift toward interoperable AI systems. The trio is transferring ownership of critical agentic technologies to establish industry-wide open standards.techbuzz+1​

The foundation immediately gains control of three essential tools. Anthropic's Model Context Protocol (MCP) enables agents to connect and interact with each other; OpenAI's Agents.md allows websites to specify rules for coding agents; and Block's Goose provides a framework that taps multiple language models to perform computer tasks.wired+1​

Nick Cooper, who oversees MCP development at OpenAI, emphasized the strategic importance: "MCP is utilized by numerous companies, yet there are others that haven't adopted it. Establishing MCP as an open standard will motivate developers and organizations to utilize it, thereby creating systems that integrate agentic AI. This open interoperability—that open standard—truly enables companies to communicate across different providers and agentic systems".wired​

The move signals a broader industry transition from chat-focused AI systems to autonomous agents that execute actions on behalf of users. While the standards remain technologically neutral, if these tools gain global prominence, they could bestow significant influence upon the U.S. companies behind them, similar to how ICANN and W3C have shaped the web's evolution.wired​

Enterprise AI Spending Explodes to $37 Billion

Enterprise generative AI spending surged to $37 billion in 2025, up 3.2x from $11.5 billion in 2024, according to Menlo Ventures' latest market analysis. This represents more than 6% of the entire software market, all achieved within three years of ChatGPT's launch.ft​

The largest share—$19 billion—went to user-facing applications and software leveraging underlying AI models. By Menlo's count, at least 10 products are now generating over $1 billion in annual recurring revenue, with 50+ products exceeding $100 million.ft​

Coding emerges as the killer use case. Departmental AI spending hit $7.3 billion in 2025, up 4.1x year-over-year, with coding capturing $4 billion (55% of the category). Fifty percent of developers now use AI coding tools daily, with top-quartile organizations reaching 65%. Teams report 15%+ velocity gains across the software development lifecycle.ft​

Enterprises prefer buying over building. Seventy-six percent of AI use cases are now purchased rather than built internally, up from 53% last year. AI deals convert to production at nearly twice the rate of traditional SaaS—47% versus 25%—reflecting strong buyer commitment and immediate value delivery.ft​

Product-led growth dominates. Twenty-seven percent of all AI application spend comes through product-led growth motions, nearly 4x the rate in traditional software at 7%. When accounting for "shadow AI adoption"—employees using personal credit cards for tools like ChatGPT Plus—PLG-driven tools may represent close to 40% of application spend.ft​

Vertical AI solutions captured $3.5 billion, with healthcare alone accounting for nearly half at approximately $1.5 billion. The ambient scribe market reached $600 million, minting two new unicorns and reducing administrative burden for physicians.ft​

Anthropic Overtakes OpenAI for Enterprise Dominance

The foundation model landscape shifted decisively in 2025. Anthropic captured 40% of enterprise LLM spending, up from 24% last year and 12% in 2023. Over the same period, OpenAI dropped from 50% to 27%, while Google surged from 7% to 21%. Together, these three companies account for 88% of enterprise LLM API usage.ft​

Anthropic's ascent was driven by remarkable dominance in coding, where it now commands an estimated 54% market share compared to 21% for OpenAI. The company has maintained an almost unparalleled 18-month run atop LLM leaderboards for coding, beginning with Claude Sonnet 3.5's release in June 2024. Despite Google's Gemini 3 Pro performance in most evaluations, it still trailed Claude Sonnet on code benchmarks, and Anthropic widened the gap again with Claude Opus 4.5.ft​

Open-source models struggle in enterprise despite broader momentum. Llama's stagnation—with no major releases since April—contributed to enterprise open-source share declining from 19% to 11%. Chinese open-source models, despite impressive progress, account for just 1% of enterprise LLM API usage.ft​

Startups Capture Two Dollars for Every One Earned by Incumbents

AI-native startups have decisively outpaced incumbents at the application layer, capturing 63% of the market (approximately $2 in revenue for every $1 earned by incumbents), up from 36% last year. This shift defies conventional expectations—incumbents possess entrenched distribution, data moats, and scaled sales teams, yet startups are out-executing much larger competitors.ft​

The dynamic varies sharply by function. Startups command 71% market share in product and engineering (code generation); 78% in sales; and a dominant 91% in finance and operations, where regulatory complexity once protected incumbents. However, incumbents retain 56% of the infrastructure market as AI app builders continue building on trusted data platforms.ft​

The AI Future: Inequality and Efficiency

Looking ahead to 2030, experts diverge sharply on AI's societal impact. The Financial Times and MIT Technology Review's "State of AI" dialogue highlights a critical concern: productivity gains will concentrate among those who can afford expensive compute.ft​

Tim Bradshaw, global tech correspondent for the FT, predicts that while robotaxis and humanoid robots will proliferate in major cities by decade's end, computational costs will render them luxuries for the affluent. Those who can afford layered AI services will enjoy compounding productivity advantages, while pricing mechanisms will exclude others. The same principle applies to physical AI—Waymo's Uber-like pricing and affordable Chinese robots create an illusion of accessibility, but the underlying computational costs will likely transform them into tools of inequality.ft​

Will Douglas Heaven, senior AI editor at MIT Technology Review, counters that rapid technological advancement doesn't immediately translate into economic or societal change—there's simply too much complicated human behavior interwoven. He notes that AI adoption in the global south lags the global north due to unreliable electricity and internet access, further widening disparities.ft​