The AI Brief

SpaceX set its IPO price at $135 per share, aiming to raise $75 billion at a $1.75 trillion valuation, a deal set to be the largest-ever IPO, surpassing the 2019 listing of Saudi Aramco. SpaceX said it plans to sell 555.6 million shares, with underwriters holding an option to purchase an additional 83.33 million shares at the IPO price, amounting to $11.2 billion. Trading opens on Nasdaq under the ticker SPCX on Friday, June 12.

The S-1 organizes SpaceX into three segments: Connectivity (Starlink), Space (Falcon 9 launches, Starship development, Starshield defense), and AI (xAI, Grok, X, formerly known as Twitter, and data centers). Starlink contributed $11.4 billion in revenue with $4.4 billion in operating profit in 2025, while the xAI segment recorded a $6.36 billion operating loss. Morningstar's fair-value range for the company is reported at $600–$800 billion based on Starlink alone, well below the IPO target.

MSCI confirmed it will apply its early large-IPO inclusion methodology to SPCX beginning on the first business day following the scheduled listing date. With an estimated valuation of approximately $1.75 trillion, SPCX would be among the ten largest constituent securities in the MSCI World and MSCI ACWI indices. All passive funds tracking those index families would be required to buy and hold SPCX shares to maintain their index-tracking weights, creating post-IPO demand that is mechanically generated and price-insensitive. Musk will own over 82% voting control after the offering.

At that size, the SpaceX IPO would dwarf Saudi Aramco's $29 billion 2019 flotation and become the largest stock-market debut ever. The scale is already fueling debate about how such a large deal could reshape capital flows across equities and digital assets. MSCI warned in a February scenario analysis that megacap IPOs expected in 2026 could unleash billions of dollars in passive investment flows, trigger sector reallocations across benchmark indexes, and drain liquidity from markets outside the newly listed companies.

The bull case rests on Starlink's 2025 revenue of $18.67 billion and its status as SpaceX's only profitable segment, alongside MSCI's confirmed early inclusion rules that could drive passive demand after listing. The bear case is equally real: SpaceX posted a $4.94 billion net loss in 2025, and Morningstar values the company at $780 billion, far below the IPO target. OpenAI and Anthropic, which have both filed S-1s confidentially, are set to be the second and third mega-IPOs following SpaceX's trillion-dollar offering.

As of May 2026, more than 80% of code merged into Anthropic's production codebase was authored by Claude, up from low single digits before February 2025. Anthropic's typical engineer now merges roughly eight times as much code per day as in 2024. Anthropic acknowledged this metric overstates the true productivity gain; an internal poll placed the median self-reported uplift at approximately 4x.

In May 2025, Claude Opus 4 averaged approximately a 3x speedup over starting code on a standard optimization task. By April 2026, Claude Mythos Preview was achieving approximately 52x. For calibration, a skilled human researcher would need four to eight hours to reach 4x. External benchmarks corroborate the trajectory: the length of tasks AI models can reliably complete autonomously has been doubling roughly every four months.

Anthropic called the threshold it is approaching "recursive self-improvement" and said the world needs a verifiable international system to slow or temporarily pause frontier AI development before that threshold is reached. The report stops short of a unilateral pledge. Anthropic said it would slow down or pause only if other frontier labs did so under verifiable conditions. Critics note the call for a pause came just days after Anthropic confidentially filed for an IPO and not long after a funding round that valued the company near $1 trillion. To skeptics, such pronouncements can read as business strategy, a way to draw regulatory scrutiny to the frontier while Anthropic continues racing toward it.

Disclosure: Anthropic, mentioned in this item, is the company that develops Claude, which generates this brief.

OpenAI's threat intelligence team tracked two distinct clusters of activity from groups with ties to China. The first, dubbed "Data Center Bandwagon," used ChatGPT to create imagery and social media comments claiming data center buildouts were raising electricity prices for Americans. The other used the tool to develop images and posts characterizing tariffs as a covert means for countries to exert control over the global technology landscape.

OpenAI traced "Data Center Bandwagon" to an unnamed Chinese technology company that does government work; that group asked ChatGPT to create comic strips about power grid capacity and electricity prices, which were later posted to X via likely inauthentic accounts alongside links to legitimate news stories. OpenAI rated the campaigns a 1 and 2 on the Bookings breakout scale, scores that indicate activity on one or more platforms but no evidence of meaningful engagement by targeted audiences.

The timing is convenient for OpenAI. The company is pushing aggressively for data center construction to meet surging demand for its products and has argued that AI infrastructure is a matter of national competitiveness with China. Framing domestic opposition as partly foreign-driven serves that agenda, even if the influence campaign identified was small and ineffective. Opposition to the construction of data centers has been on the rise in the U.S., with at least 36 projects blocked or delayed between May 2024 and June 2025, according to Data Center Watch.

Enterprises often want to deploy AI through the procurement processes and governance frameworks they already trust. To help make that happen, OpenAI and Oracle are partnering to make OpenAI frontier models and Codex easier to access for Oracle Cloud Infrastructure customers. In the coming weeks, Oracle customers will be able to apply eligible Oracle Universal Credits toward OpenAI models and Codex through OCI, giving customers a path to access OpenAI models under their existing purchasing workflow and cloud commitment.

The pattern mirrors OpenAI's earlier integrations with Microsoft Azure and Amazon Bedrock, both of which moved AI adoption from standalone procurement into existing hyperscaler relationships. Enterprises often want to deploy AI through the procurement processes and governance frameworks they already trust. Oracle's customer base skews toward large regulated industries, including financial services, healthcare, and government, where new vendor onboarding cycles can run six to eighteen months. Folding OpenAI access into an existing cloud commitment bypasses that cycle. No pricing terms were disclosed; Oracle customers were directed to their sales representatives for availability details.