To The Moon

By now, I am certain readers would have heard about Anthropic’s leaked IPO prospectus.

The one with information that jolted the industry - generating both excitement over a potentially market-defining listing while provoking alarm about the financial & existential risks exposed by the filing.

Anthropic’s IPO case

Reuters reportedly reviewed the confidential document, while no official public filing had yet appeared.

Here’s the skinny.

• Anthropic’s FY 2025 revenue rose twelvefold, or +1,088%, to $4.59 billion.

• However, the company also recorded an $8.06 billion operating loss and a $41.97 billion net loss, compared with an $8.31 billion net loss in 2024.

• Approximately $34 billion of the 2025 net loss reportedly reflected an accounting charge linked to the higher estimated value of financing that could later convert into shares.

• Anthropic has spent $7.33 billion on compute and infrastructure in 2025, 3x its 2024 spending and more than ½ of total operating expenses of $12.65 billion.

• More importantly, the prospectus reportedly outlines $518 billion in future cloud, computing, and infrastructure obligations.

The figure highlights both - (a) scale of Anthropic’s growth strategy and (b) extent to which the company’s success is tied to suppliers, cloud providers, chipmakers, lenders, and other parts of the AI ecosystem.

Anthropic is reportedly targeting a valuation above $2 trillion, more than double its estimated $965 billion valuation mentioned in May 2026 and above $SpaceX(SPCX)$ reported $1.77 trillion valuation after its June 2026 IPO.

The offering could come after the November 2026 US midterm elections, and early investors may be allowed to sell shares.

The prospectus may therefore become an important valuation benchmark for other AI companies, including OpenAI.

The company’s reported Q2 2026 performance strengthens the growth argument.

The Compare.

Anthropic generated $11.5 billion in revenue, up fourteenfold YoY, compared with OpenAI’s $6.7 billion.

It also reported a $559 million adjusted operating profit, while OpenAI recorded a $12.3 billion operating loss.

Product updates to Claude Code and Claude Cowork were associated with sharply higher sign-ups and growing investor expectations that AI applications could replace parts of traditional software.

The Risks.

Anthropic’s filing also contains unusually serious safety disclosures.

Risk factors reportedly occupy about 80 of the prospectus’s 261 main-body pages, compared with 48 pages describing the business.

The company warned that advanced AI could :

• Pose existential risks to humanity.

• Its AI models might r(a) esist shutdown, (b) conceal or manipulate information, (c) behave in ways resembling blackmail, or (d) recognize when they were being evaluated.

Anthropic further added :

• Such model awareness could limit its ability to assess safety.

• That the safety work used about 6% of AI research compute during a sample week in July 2026.

CEO Dario Amodei reportedly issued a similar warning to the United Nations Security Council.

Honestly, is Anthropic trying to ‘cover its ass’ in the course of IPO so that it could say ‘I told you so’ in the event of a major security disaster like the activation of a nuclear war, depicted in many Hollywood movies?

The central IPO question is whether Anthropic can convert explosive revenue growth into sustainable profits before its infrastructure commitments become a financial burden.

Top 10 AI stocks

The following ranking uses September 2026 market-cap data and treats “AI stocks” broadly, including companies whose businesses depend heavily on AI chips, cloud infrastructure, AI platforms, or AI-enabled products.

Market capitalizations change continuously. As such, different providers may produce slightly different figures.

Therefore, rankings should be read as approximate rather than fixed values.

The significance of Anthropic’s leaked IPO extends beyond the company itself - its valuation, net loss, and future infrastructure obligations - providing a direct lens through which investors can reassess its growth prospects, capital demands, and risks facing the 5 largest AI-linked stocks.

The late-night Reuters leak of Anthropic’s confidential IPO prospectus on Mon, 28 Sep 2026, sent shockwaves through the technology sector.

The leaked document revealed a staggering $42 billion net loss for 2025 alongside an eye-popping $518 billion in future cloud, computing, and infrastructure commitments.

This revelation served as a massive "AI gut-check" on Wall Street.

While the $2 trillion target valuation underscored massive industry optimism, the scale of the losses and structural warnings that advanced AI poses "existential risks to humanity" caused US stock futures to waver and tempered enthusiasm for mega-cap tech.

The 10 stocks previously listed were affected through 2 primary market reactions:

1. The Infrastructure Backbone (The Winners)

Anthropic's disclosure that it plans to spend over ½ a trillion dollars on cloud computing and hardware directly validates the revenue pipelines of major infrastructure providers.

Amazon (AMZN) & Alphabet (GOOGL):

• As Anthropic’s early strategic partners and primary cloud providers, their infrastructure demand was heavily validated.

• The leak explicitly highlighted massive multi-billion-dollar infrastructure contracts signed with both companies.

• This included a reported $100 billion cloud deal with AMZN and a $200 billion pact with GOOG.

• And provided strong fundamental support for both stocks against broader market dips.

Nvidia (NVDA), AMD (AMD) & Broadcom (AVGO):

• Hardware and networking providers saw structural validation.

• Anthropic alone spent $7.33 billion on compute and data center infrastructure in 2025, a threefold surge from 2024.

• Knowing that frontier labs must maintain this capital expenditure (capex) to stay competitive keeps demand narrative intact for chipmakers.

2. The Commercial Valuation & Safety Shockwave (The Risks)

Conversely, the operational metrics and risk warnings introduced immediate macro headwinds for software and direct model competitors.

Microsoft (MSFT):

• Faces indirect pressure as Anthropic's numbers cast a harsh light on true cash-burn rate of frontier model development.

• With OpenAI (MSFT's key partner) also prepping a confidential IPO filing, Wall Street began aggressively recalculating the heavy toll that OpenAI's demand backlog takes on MSFT’s free cash flow.

Meta (META):

• Felt the friction of safety concerns.

• The prospectus noted that autonomous models could act deceptively (e.g., "resisting shutdown" or "manipulating information").

• Combined with news that OpenAI simultaneously shelved a model over similar safety anomalies, it triggered broader regulatory anxieties

Palantir (PLTR) & Software Applications:

• Although not a Top 10 AI-stock, the impact could be far reaching.

• Pure-play AI enterprise platforms faced market skepticism over the long-term sustainability of AI deployment costs.

• Critics questioned whether massive infrastructure obligations can ever realistically translate into profitable software revenue.

My viewpoints: (mine only)

Anthropic’s proposed IPO would force investors to decide whether explosive AI growth justifies in one valuation:

• Extraordinary losses.

• Infrastructure commitments.

• Last but not least, safety risks.

The most important signal may not be whether Anthropic reaches $2 trillion, but whether public markets reward a business that openly acknowledges that its ‘flawed’ technology could threaten humanity.

If financial ambition continues to outrun technical control, the AI boom may be tested not by a shortage of demand, but by the consequences of achieving the wrong kind of success.

In a weak attempt to cull concerns, Trump’s Tue, 29 Sep 2026, White House AI luncheon sought to balance rapid innovation and US competitiveness with concerns over AI safety and oversight.

Participating technology companies signed a voluntary “White House Accord on Superintelligence” covering (1) internal safeguards, (2) monitoring, (3) independent audits, and (4) continued cooperation.

However, the meeting failed to produce (a) binding regulations, (b) enforcement mechanism, or (c) development moratorium, leaving the AI-companies largely responsible for policing the risks themselves.

We already know how effective “self-policing” by tech giant themselves, don’t we.

Just look at Mark Zuckerberg who has testified before US Congress at least 8 times and specifically appeared and apologized during major high-profile Senate hearings on:

• 2007 - Beacon Advertising disaster.

• 2010 - The "Dumb F***s" Instant Messages leaked .

• 2011 - The Federal Trade Commission (FTC) settlement on consumers deception.

• 2017 - Virtual Reality Tour of Devastated Puerto Rico to showcase their virtual reality app.

• 10 Apr 2018 - Cambridge Analytica scandal

• 31 Jan 2024 - Child Online Safety hearing

Do we forsake humanity Just to make a dollar profit, just think about it for one second. Now, what do you think ?

Remember to check out my other posts. (See below). Help to Repost ok, Thanks.

Must Read: Click on below titles to access. Repost to share, Like as encouragement ok. Thanks.

• VOO's Historic gains, Risks to investors. Really?

• US Market - Week that Was & Will Be.

• Wed (23/09), start of US Market pullback?

• Do you think AI really possess a threat to humanity is real ?

• Do you think Trump’s white house effort (or lack thereof) to rein in rogue-AI harms via a “self policing” stance sufficient to address real world risks & concerns?

If you find this post interesting, give it wings! ️ Repost and share the insights ?

Do consider “Follow me” and get firsthand read of my daily new post. Thank you.

@Daily_Discussion

@TigerPM

@TigerStars

@Tiger_SG

@TigerEvents