anthropic

Google (GOOG.O) collaborates directly with Anthropic on TPU and partners with Broadcom (AVGO.O).

According to Bloomberg, Anthropic warned in its IPO-related documents that its public advocacy for AI safety regulation and statements from executives may expose the company to risks such as pressure from the Trump administration and selective scrutiny. This risk is related to the company's multiple conflicts with the Trump administration, including disputes over requests to impose safety restrictions on the use of its AI technology by the Department of Defense, as well as disagreements regarding the safety of the Mythos model and export controls.Anthropic also warned that even if other similarly situated companies are not treated equally, Anthropic may still become a specific target for certain departments of the Trump administration. Future export restrictions and cybersecurity or biosafety incidents triggered by AI models could impact its product deployment, international expansion, and business growth.The company also acknowledged that once the regulatory measures it advocates are implemented, they may increase compliance costs, restrict product development and release, and slow its own growth.

According to TNW, Anthropic announced the acquisition of Mithril Medicine in a blog post on Tuesday.This company has only one employee and is developing therapies for two rare muscular dystrophies. The terms of the deal were not disclosed, and founder Cody Tranbarger has joined Anthropic.

The Hang Seng Index Company announced that the Hang Seng Hong Kong American Technology Index will add Anthropic, PBC (stock code: ANT.O) to the list of designated U.S. listed constituent stocks, provided that Anthropic confirms its listing. The weight of the 9 U.S. listed constituent stocks will be recalculated to equal weight after the close of trading on the 7th trading day following Anthropic's listing, with the new constituents and new weights taking effect on the 11th trading day.The index will also update the stock selection criteria, constituent stock selection methods, rapid inclusion mechanisms, and replacement rules for temporarily removing constituent stocks, and will introduce new liquidity requirements. The new methods will be adopted starting from the index review ending on September 30, 2026, with the relevant constituent stock changes taking effect on December 14, and the index name and code will remain unchanged.

The U.S. Department of Justice has filed an appeal, attempting to reinstate a nationwide ban against the artificial intelligence company Anthropic. The next day, the General Services Administration (GSA), responsible for government procurement, announced an expanded collaboration, allowing government agencies to use Claude at a 40% discount, paying based on actual usage without a commitment to a minimum expenditure. Anthropic disclosed that approximately 500 federal, state, and local government organizations in the U.S. are currently using Claude, with usage by federal civilian agencies increasing about 20 times over the past year.The GSA publicly supported Trump's ban on Anthropic in February of this year and removed Claude from the government procurement list. In March, a California court halted the comprehensive ban, and the GSA restored procurement eligibility in April. In August, the court further ruled that the government's broad ban was illegal. Although the Department of Justice has appealed, the relevant ruling will not automatically become invalid as a result.

Anthropic has selected JPMorgan Chase to handle its IPO stock sales aimed at retail investors in the United States, expecting to allocate about 15% to 20% of the shares to retail investors.JPMorgan Chase CEO Jamie Dimon will participate in the IPO activities. Citigroup is responsible for coordinating the sale to international retail investors, and the IPO will also take place in Australia, Canada, Japan, some European Economic Area countries, Switzerland, and the United Kingdom.

According to BeInCrypto, the U.S. government sold 4.45 million shares of Anthropic, cashing in $627 million to repay FTX creditors.According to Anthropic's S-1 filing, these shares were resold to 9 buyers. Based on the valuation before Anthropic is set to go public on Nasdaq under the ticker ANT, this batch of shares is currently valued at approximately $3.94 billion.

The U.S. Department of Justice submitted a notice of appeal to the Ninth Circuit Court of Appeals, attempting to overturn the ruling made by a California court in August, which restored a full federal agency ban against the developer of Claude, Anthropic. On October 26, local time, Anthropic publicly submitted its IPO prospectus, and on the same day, the U.S. Department of Justice filed the aforementioned appeal.In September, the Washington appellate court allowed the Pentagon to exclude Claude from military systems and related defense procurement on the grounds of supply chain risks. The California court had previously blocked a broader ban that not only required all federal agencies to stop using Claude but also prohibited defense contractors from engaging in non-military commercial cooperation with Anthropic. The judge believed that the government lacked sufficient evidence to prove the associated risks, and the ban was essentially retaliation against Anthropic for publicly criticizing the government.The prospectus shows that government agency contract revenue accounts for only about 1% of Anthropic's total revenue. Evidence cited by the California court indicated that three government contractors terminated their contracts with Anthropic or were asked to cease cooperation, and three deals worth over $180 million, which were close to being finalized, fell through. Some clients requested the addition of unilateral termination clauses, while others turned to competitors. Anthropic had previously estimated to the court that if the broadest ban were fully implemented, revenue could decrease by billions of dollars by 2026.

The AI laboratory Anthropic, which developed Claude, named TSMC in its IPO prospectus, warning that the world's most advanced chip manufacturing capacity is highly concentrated in Taiwan. The prospectus states that if TSMC's operations are disrupted, it could severely impact the supply of AI chips and affect Anthropic's business, financial condition, and operational results.According to MoneyDJ citing the prospectus, the global AI industry heavily relies on TSMC's advanced processes. Suppliers such as NVIDIA and AMD, as well as companies that design their own chips like Amazon and Google, all need TSMC to turn their designs into physical chips. The listed risks include earthquakes, typhoons, interruptions in power and water supply, as well as cross-strait geopolitical tensions, military conflicts, blockades, tariffs, and trade restrictions.The prospectus also states that if a significant event occurs in Taiwan, the supply of advanced chips could be severely hindered; due to the time-consuming nature of building and certifying alternative foundries, the impact could last for several years. The expansion of advanced processes outside Taiwan is still in its early stages, making it difficult to effectively reduce concentration risks in the short term. TSMC has been unable to fully meet market demand due to strong demand, resulting in Anthropic not being able to obtain a sufficient supply of chips in a timely manner.

The founder and trader of the NFT project DeGods, Frank, posted on X that although Anthropic's latest S-1 shows strong revenue growth, it cannot be concluded that the resistance to AI investment has disappeared. He stated that the second quarter of 2026 will be a peak for companies to purchase and consume AI Tokens in large quantities, and the market will only begin to pay attention to AI spending management by July; the rise in model homogeneity and the narrowing of output differences may intensify competition and compress profit margins, a trend that has not yet fully manifested in the second quarter.Frank also mentioned that there is still a lack of clear evidence proving that AI can create more profits for most companies, and Anthropic's large enterprise clients are also under pressure to improve their profit margins, which may lead to cuts in spending related to Claude in the future. Large tech companies and AI labs have signed capital expenditure commitments extending to 2034, and related expectations may have already been factored into stock prices; if companies improve the efficiency of existing chips to reduce costs or issue more shares to raise additional capital expenditure, it could put pressure on the valuations of related stocks. He expressed a long-term optimistic view on AI and may participate significantly in the Anthropic IPO, but believes that the ultimate winners in the industry have yet to be determined, and real investment opportunities may only emerge after the current market cycle ends.

According to Bloomberg, the Commonwealth Bank of Australia may profit from Anthropic's initial public offering. The bank participated in Anthropic's Series E funding round in March 2025, with its holdings valued at 153 million Australian dollars, approximately 107.5 million US dollars, as of June 30, 2025.By the end of June 2026, as Anthropic's valuation rose to 965 billion US dollars, the bank's holdings had increased in value to 1.53 billion Australian dollars, approximately ten times higher than before. Bloomberg stated that this valuation increase is considered paper gains, and whether it can ultimately be realized through the IPO depends on the listing process and related transaction arrangements.

After Anthropic publicly submitted its prospectus the previous evening, investment institutions and market participants began to assess its growth quality, cost structure, and potential valuation. Morningstar senior equity analyst Malik Ahmed Khan and PitchBook analyst Rolfes stated that as of September 30, Anthropic had over 11,900 enterprise customers contributing at least $100,000 in revenue over the last 12 months, compared to about 950 in the same period last year. Khan believes this reflects a faster adoption of Claude in the enterprise market and significant revenue growth.Analysts also pointed out that the rapid increase in customer numbers alone is insufficient to indicate long-term revenue stability. The prospectus lacks detailed data on customer retention, and the disclosed net revenue retention rate is also affected by a small base; as the comparison base expands, this metric may decline. Investment in computing power is another focus. Morningstar believes that Anthropic currently demonstrates a certain level of computing efficiency and operational efficiency, but some computing resources and related expenditures have not yet been fully reflected; as more computing power is utilized, costs may rise, and if revenue growth does not keep pace, profit margins will be under pressure.PitchBook analyst Rolfes reminded that the scale of future computing contracts and infrastructure commitments is enormous, and whether revenue growth can be sustained is an important variable in assessing profitability prospects. Gavin Baker, managing partner and chief investment officer of Atreides Management, stated that it is still possible for Anthropic to reach an annualized recurring revenue of $100 billion by the end of the year; this statement reflects his personal judgment and is not the company's published performance guidance. Nigel Green, CEO of independent financial advisory firm deVere Group, questioned the market's rumored potential valuation of $2 trillion, calling this expectation "absurd," and reminded to pay attention to infrastructure investments, slowing growth, and future financing needs. The $2 trillion figure remains a market rumor, and the prospectus has yet to determine the offering price and scale.

Anthropic has officially submitted its IPO application. According to its S-1 cash flow statement disclosed in the prospectus, the company has raised approximately $140 billion in cash through preferred stock, convertible bonds, and SAFE agreements from 2023 to the first half of 2026, of which $110 billion was raised in the first half of 2026, including $65.2 billion in Series H financing.As of June 30, 2026, Anthropic's remaining cash includes $112.4 billion in cash and cash equivalents and $19.3 billion in short-term investments, totaling approximately $132 billion, along with an additional $2.5 billion in restricted cash. Other financing channels include a $15 billion undrawn revolving credit facility expanded in September 2026, and up to $77 billion in partner note financing, of which $20 billion from Google, $15 billion from Amazon linked to computing milestones, and $42 billion from Broadcom solely for lease payments, with IPO proceeds to be added on top of this.Accordingly, Anthropic's cash holdings are higher than those of tech giants like Amazon, Meta, and Microsoft, second only to Google's parent company Alphabet's $242 billion. Its projected expenditure report also shows that starting from December 2026, it has committed approximately $37 billion annually for computing power solely to SpaceXAI, with total computing power commitments estimated at around $780 billion to $810 billion.

According to a report by Financial Associated Press, based on the IPO filing released by Anthropic on Monday, sales generated from the cloud markets of the three major cloud service providers—Google, Amazon, and Microsoft—accounted for 48% of Anthropic's total revenue in the first half of 2026, up from 11% in 2023. These cloud service providers collect payments from customers and transfer them to Anthropic after deducting fees, with payments to channel partners increasing by $2.7 billion year-on-year in the first half of this year. The document shows that Anthropic's sales in the second quarter were $11.555 billion, with an operating profit of $974 million.The S-1 describes the aforementioned cloud service providers as simultaneously acting as investors, customers, distribution partners, and competitors, stating that they can understand Anthropic's pricing and commercial terms. It also mentions that these companies may prioritize promoting their own AI models, raising antitrust concerns. The document lists the minimum commitments for computing power as follows: at least $114.5 billion to Google by 2033, at least $110 billion to Amazon by 2036, and $44.9 billion to Microsoft; the minimum commitments remain valid regardless of whether the computing power is actually used. Additionally, it is expected that between 2028 and 2031, $125 billion to $150 billion will be paid for additional TPU computing power.In terms of equity, Amazon and Google currently hold shares without voting rights and are not included in the list of securities with more than 5% voting rights; after the IPO conversion, both will exceed 5% ownership of preferred shares and certain convertible notes. Amazon has invested approximately $18 billion to date and has committed to invest up to an additional $15 billion under specific conditions, with related notes convertible into a maximum of 1.84 billion shares, equivalent to 10.9% of fully diluted shares as of June this year, and subject to a 12% ownership cap. Google has invested approximately $23 billion and has committed to invest an additional $20 billion upon meeting certain conditions, with the latest agreement having a maximum ownership cap of 14.99%. The document also lists a $161.2 billion equipment leasing agreement related to Broadcom, as well as a computing power procurement arrangement with SpaceXAI for up to $120.5 billion.

According to The Washington Post, AI cybersecurity startup Cathedral, founded by former members of the U.S. "Department of Government Efficiency" (DOGE), claims that its AI agent breached Anthropic's internal systems in less than 12 hours, obtaining content such as meeting notes, salary and hiring information, and Slack messages. Anthropic confirmed that the intruder accessed parts of its sales, human resources, and IT systems, involving business contact information of clients and partners, as well as some basic user account records, but did not involve Claude user conversations, source code, or model weights.This test was conducted under Anthropic's vulnerability bounty program, with the intrusion achieved through vulnerabilities in third-party commercial software; Anthropic accused Cathedral of not fully disclosing its actions, violating program rules. Cathedral is valued at $1.4 billion and has secured $160 million in funding from a16z, Sequoia Capital, and others.

According to The Defiant, after Anthropic submitted its IPO documents, the cryptocurrency exchange OKX launched a 10x betting product based on Anthropic's valuation.This product is aimed at eligible traders in the European Economic Area, offering cash-settled risk exposure rather than actual shares or IPO allocations. The product includes a share quantity adjustment mechanism designed to maintain the value of the position.