Anthropic PreStocks (ANTHROPIC) on Solana

Anthropic's annualized revenue stands at over $65 billion compared to OpenAI's $40 billion — a roughly $25 billion gap that has drawn scrutiny over how each company accounts for cloud partner sales. Anthropic records the full value of Claude API sales through cloud providers like AWS as top-line revenue, listing provider fees as expenses, on the basis that it is the principal in those transactions. OpenAI, by contrast, records only its net share from partner arrangements including Microsoft.

Accounting professor Francine McKenna notes that both treatments can be defensible depending on customer contracts and how each company justifies its principal-versus-agent classification to the SEC. Critically, even if Anthropic adopted OpenAI's more conservative net-revenue approach, the adjustment would reduce its reported figures by only 6–10%, still leaving Anthropic roughly $19–21 billion ahead. Anthropic's forthcoming IPO S-1 is expected to formally document its revenue recognition methodology, giving public investors a clearer picture of the underlying economics.

Anthropic is overhauling how it handles enterprise customer data after significant pushback against a mandatory 30-day retention requirement introduced in June 2026 alongside the Claude Fable 5 and Mythos 5 model launches. The original policy required all traffic on those models to be retained by Anthropic for 30 days to defend against misuse and novel cyberattacks — with a pledge that the data would not be used for model training or non-safety purposes — but enterprise customers raised concerns over where the data sat and who ultimately controlled it.

In response, Anthropic is rolling out a new program called Enterprise Frontier Safeguards, which shifts data custody from Anthropic's own infrastructure to the customer's cloud environment. The 30-day retention window remains, but under the new arrangement companies manage review and storage within their own systems. The phased rollout is targeting broader availability in fall 2026. The concession signals that Anthropic's path to winning large enterprise contracts — and sustaining the revenue growth it has showcased ahead of its IPO — runs directly through meeting corporate data governance standards, even when those requirements complicate the safety architecture it has built around its frontier models.

Anthropic has agreed to a \$35 billion cloud computing contract with Lambda, an Nvidia-backed cloud provider, to secure additional GPU capacity for its Claude AI products. The infrastructure is centered on a data center in Nueces County, Texas, developed by Hut 8, delivering approximately 350MW of computing power via Nvidia-powered systems. The deal is designed to bring more capacity online specifically to support demand for Claude and associated tools including Claude Code.

The agreement follows Anthropic's \$45 billion commitment to Nscale's West Virginia data center campus announced the prior week, bringing the company's recently disclosed cloud infrastructure commitments to \$80 billion. The scale and pace of these deals signal that Anthropic is aggressively locking in compute supply ahead of its anticipated IPO, betting that demand for Claude-powered services will require substantial dedicated infrastructure rather than reliance on existing hyperscaler allocations.

Sony Music Publishing and Warner Chappell Music have filed a copyright lawsuit against Anthropic, alleging the company conducted a "brazen campaign of illegally torrenting, scraping and downloading copyrighted works on a massive scale" to train its Claude AI models. The publishers describe the conduct as "one of the largest and most blatant ongoing thefts of intellectual property in history," claiming Anthropic harvested thousands of copyrighted musical compositions without authorization and stripped copyright management information from the works. Statutory damages could reach $150,000 per infringed work and $25,000 per instance of copyright management information removal, with the scale of alleged violations pointing toward potential aggregate exposure in the billions.

This is the third major copyright action filed against Anthropic, following a suit by Concord Music Group and Universal Music Group seeking $3 billion and a separate author settlement reportedly valued at $1.5 billion. The accumulating litigation presents a material disclosure challenge as Anthropic advances its IPO preparations: prospective public investors will need to weigh open-ended copyright liability against the company's strong revenue trajectory. How Anthropic frames and reserves against these claims in its S-1 filing is likely to be a focal point for institutional underwriters pricing the offering.

Anthropic's central pitch to prospective IPO investors is that it has addressed AI's most critical constraint: compute. The company's $45 billion, six-year agreement with Nscale — combined with existing arrangements with Amazon, Google, and SpaceX — is designed to demonstrate that infrastructure bottlenecks will not impede its growth trajectory ahead of an expected October 2026 IPO. Anthropic filed confidentially in June and is now in active investor meetings, targeting a valuation of $2 trillion or more, which would rank among the largest IPOs in history. The company is pitching a $30 trillion total addressable market, a figure finance analysts have received with skepticism.

The financial metrics framing that case are substantial: Anthropic reported Q2 2026 revenue of $11.6 billion — surpassing OpenAI for the first time on a quarterly basis — with an annualized run rate reaching approximately $65 billion by July. Its valuation has climbed sharply, from $965 billion in May to the current $2 trillion target. Analysts flag several risks: the 460-megawatt Nscale facility is not scheduled to come online until late 2027, profit margins face pressure under heavy infrastructure obligations, and Anthropic's current share of its stated addressable market remains below 1%.

A California federal judge ruled on August 28 that the Pentagon's designation of Anthropic as a national security supply chain risk was unlawful, finding it violated the First Amendment and was "arbitrary and capricious." Judge Rita Lin concluded the Defense Department's blacklisting — which effectively barred Anthropic from government work — constituted "unlawful retaliation" aimed at making "a public example" of the company for criticizing administration policy. She also noted an inherent contradiction in the government's position: Defense Secretary Pete Hegseth had previously threatened to invoke the Defense Production Act to compel Anthropic's cooperation, implying the company was "essential to national security rather than a threat to it."

The dispute originated in early 2026, when Anthropic refused Pentagon demands to remove safeguards from its Claude AI systems that would have permitted "any lawful use," including fully autonomous weapons and mass domestic surveillance. CEO Dario Amodei stated in February that the company "cannot in good conscience" strip those protections, prompting the March 2026 supply chain designation. Anthropic welcomed the ruling, saying it remains "focused on working productively with the government to harness A.I. for our national security." The decision removes a significant regulatory overhang for a company that filed for an IPO expected to be one of the largest in U.S. history.

Anthropic is preparing for what could be the largest IPO on record, with the AI company valued at $965 billion following a $65 billion raise in May 2026 and now targeting more than $75 billion in its public offering — a figure that would exceed SpaceX's recent debut. The company was founded in 2021 by former OpenAI executives Dario Amodei (CEO) and Daniela Amodei (president) and employs roughly 5,000 people. Its annual revenue run rate has surpassed $65 billion, fueled by strong demand for developer-focused products including Claude Code, and Anthropic has emerged as the leading rival to OpenAI by pursuing a narrower strategy centered on coding and enterprise tools.

Despite the revenue trajectory, Anthropic lost approximately $42 billion in 2025 and is expected to continue burning cash for years. The company also faces political headwinds: an ongoing legal dispute with the Trump administration over government contracts and military AI access, combined with CEO Dario Amodei's public warnings about AI dangers and advocacy for regulation, have drawn criticism from some conservatives. Investors bullish on the $30+ trillion total addressable market the company targets will need to weigh those cash losses and political risks against Anthropic's rapid revenue growth ahead of its Wall Street debut.

Anthropic is reportedly in preliminary talks to lease a large data center in the Cedar Creek area of Bastrop County, Texas. The project centers on a 2,842-acre site anchored by a gas-fired electric generation facility purpose-built to power the data center campus. BlackChamber Partners LLC, a Washington, D.C.-based firm, and Dallas-based Pacifico Energy LLC are among the parties working with Anthropic on the project, though sources indicate no agreement has been finalized.

The potential Bastrop deal follows a separate data center lease Anthropic recently secured in neighboring Milam County, northeast of Austin, pointing to an accelerating infrastructure buildout across central Texas. As Anthropic prepares for a high-profile IPO, expanding owned and leased compute capacity is central to its ability to train increasingly large models and serve surging enterprise demand for its Claude AI systems.

With Anthropic's S-1 expected before the end of August 2026, the filing will for the first time formally disclose Amazon's exact ownership percentage — a figure that has significant implications for AWS revenue modeling. Amazon has invested $13 billion in Anthropic to date, with up to an additional $20 billion contingent on commercial milestones. Anthropic's current valuation stands at approximately $965 billion following its $65 billion Series H, and the company is targeting an IPO raise that would match or exceed SpaceX's recent record.

The Amazon-Anthropic relationship extends well beyond equity. Anthropic has pledged more than $100 billion in AWS infrastructure spending over ten years, deployed over one million Trainium chips under Project Rainier, and powers Amazon Bedrock — which serves more than 100,000 customers using Claude models. With quarterly revenue now exceeding $11.5 billion and an annualized run rate approaching $65 billion, the S-1's disclosure of the full commercial dependency between the two companies is expected to give AWS investors a clearer framework for modeling the cloud revenue acceleration tied to Anthropic's infrastructure demand.

Investors in Anthropic are reportedly pushing for a $2 trillion valuation in an expected October IPO — roughly double the $965 billion valuation the company secured when it raised $65 billion just days before filing a confidential draft registration statement on June 1. Notably, that $2 trillion figure comes from investors, not from Anthropic's senior executives, who have not privately endorsed the target. The company's financials support rapid growth — second-quarter revenue exceeded $11.5 billion, more than double Q1's $4.73 billion — with investors projecting annualized revenue of $100–120 billion by year-end.

The cautionary backdrop is SpaceX, which holds the current IPO record at a roughly $1.77 trillion valuation after raising approximately $75 billion at $135 per share in June. Despite reporting 92% revenue growth and $14.1 billion in new contracts, SpaceX shares surged 67% to $225.64 before falling 22% below the offer price to $104.83, and now trade near the original $135 — leaving initial buyers essentially flat. Analysts note the dynamic reflects a recurring pattern where offer prices are already calibrated to reflect known growth, leaving limited upside for public-market entrants even when business fundamentals remain strong.