The INTEL  

Anthropic’s latest IPO filing, Amendment No. 2 on November 10, says the company has determined it met its own safety threshold for AI-accelerated research, puts its committed spending on compute at about $900.9 billion, and adds $26.0 billion of shares to be bought by seven of its suppliers and partners. For any organization that depends on its models, this is the most detailed supplier disclosure available.

What changed since Amendment No. 1 (November 2), the filing E97 analyzed:

  • The safety threshold. Amendment No. 1 said Anthropic “may imminently meet” its Responsible Scaling Policy’s definition of “dramatically accelerated progress attributable to AI.” Amendment No. 2 says “we made a determination that we have met this threshold based on what we expect from future models and our overall rate of progress, rather than on what we have observed from current models.” It says it is evaluating the implications, will publish further reports, and is considering policy updates “to keep humans in control of frontier model development.”
  • The commitments. About $900.9 billion of committed spending, substantially all for compute and infrastructure, up from $895.4 billion a week earlier. About $451.6 billion falls before the end of 2030, and $70.7 billion within twelve months of September 30.
  • Suppliers become shareholders. Google, NVIDIA, AMD, Broadcom, Micron, Microsoft and TSMC have agreed to buy $26.0 billion of non-voting Class N shares at the IPO price, closing right after the offering.
  • The offering. An expected price range of $70 to $78 a share, on Nasdaq under the proposed symbol “ANT.”

What stayed the same: Amazon, Google and Microsoft “simultaneously serve as our investors, customers, distribution partners, and competitors,” and 44% of revenue in the first nine months of 2026 came through the big providers’ cloud marketplaces.

Disclosure: FIR Risk uses AI models from several providers, including Anthropic.

Why It Matters  

This is not investment commentary. It is supplier due diligence. When a critical process depends on one AI model, the model maker is a critical supplier, and a prospectus is the most complete public account of that supplier’s risks you will get.

Two things stand out for buyers. First, the circle is closing: the companies that sell Anthropic its chips, cloud and distribution will now also own part of it. If your AI, cloud and chip supply all trace back to the same few companies, a problem at one of them reaches you through several paths at once.

Second, the safety determination is a provider telling investors that its own policy now applies to how it develops future models. The filing says Anthropic is evaluating what meeting it means and will publish further reports. For a buyer, that is a factor in how and when new models arrive, and it is worth tracking the same way you track a supplier’s roadmap.

What To Do — One Key Action  

If a critical process depends on a frontier AI model, add the provider’s own public disclosures to your third-party review: who owns and supplies it, what it has committed to spend, and what its safety policy says happens as models become more capable. Then confirm the process could move to another model if terms or availability change.

MITRE ATT&CK  

  • Not technique-specific. This INTEL is about supplier concentration and AI provider risk.

Learn More  

Powered by FIR Risk Platform — AI-driven threat intelligence for enterprise risk leaders.