Anthropic's IPO becomes the first public mark for frontier labs, with a thin float and a trust in charge
Frontier labs compete for one pool of investors, and governance structures decide who sets the public price.
Early · 5% share pledge + OpenAI's $30B race
Written for: Investor or analyst
Next step
Thesis: Anthropic's listing sets the public reference price for frontier labs, and its trust and reduced float make that price more volatile than its revenue alone would justify. Confirming signal: the S-1's disclosure of trust powers and float size, and whether OpenAI closes its round before the listing. Already building: Anthropic and OpenAI on the issuer side, with Amazon, Google and Microsoft as the strategic holders whose stakes get repriced.
Why now
Earlier IPO cards were about revenue quality. What's new is the market structure. A smaller float plus an unusual controlling trust means a few holders set the price, and that price becomes the reference mark for every private AI round and for the hyperscalers that own lab stakes. It also lands in a fragile market: neocloud stocks just sold off, and Anthropic says AI agents log six times as many hours as its staff, a fact that will feed public arguments about jobs and governance.
What became possible
Anthropic pledged 5% of its stock to a public benefit fund set up as an LLC, taking those shares out of the tradable float, ahead of an expected mid-November IPO. OpenAI is racing to raise about $30B at a valuation above $1.4 trillion before that debut, because both labs draw on the same investors. Kalshi shows Anthropic as the last big AI IPO of 2026.
Two ways in
- Lab governance and float tracker
Structured data on trust powers, share classes, lockups and float for each frontier lab, built for institutional investors.
- Customer
- Public-market funds and index providers
- Model
- Research subscription
- Private-mark reconciliation service
A service that marks private AI holdings against the Anthropic public price, adjusted for structure and rights.
- Customer
- Venture funds and LPs holding AI lab exposure
- Model
- Per-fund valuation fee
Competition
A new entrant would win by being the defensible, documented methodology that auditors accept for OpenAI, xAI and similar holdings during the next two or three quarterly marks. That probably means partnering with or selling to the valuation firms rather than going straight to LPs. It also depends on data access, because the investor-rights agreements are not public. The window may be short, because Caplight or 73 Strings could add a public-anchor adjustment quickly once Anthropic trades.
Full competition notes (9 competitors)
I found no product that does either idea specifically for frontier labs. The pieces exist separately, though, and the incumbents next door are strong. For the governance and float tracker: FactSet SharkRepellent and ISS already sell structured charter and dual-class data to the public-market funds and index providers you'd be targeting. They will cover Anthropic's charter once the S-1 is public. As of mid-September 2026 it had only been filed confidentially, and press reports point to an October or November listing at roughly $1.5T to $2T. What they don't do is model the unusual parts. That means Class T stock and the Long-Term Benefit Trust's power to elect board seats, the conversion terms that need the trust's consent, PBC duties, and side letters with hyperscalers. They also don't apply any of this to the labs that are still private, such as OpenAI's nonprofit-controlled PBC, xAI and Mistral, where index providers and crossover funds need to plan float and inclusion before those labs list. The opening is a narrow, high-touch dataset covering the 6 to 10 labs, kept up to date from charters, S-1s and reported financings, and sold before the incumbents decide this niche deserves custom coverage. The risk is that the market is small and that FactSet or MSCI can copy the format once enough labs are public. For the reconciliation service: Caplight and Forge already supply secondary marks for Anthropic and OpenAI, and valuation software like 73 Strings plus the big valuation-advisory firms (Kroll, Houlihan Lokey) are where funds actually do their ASC 820 marks. Nobody I found yet uses a liquid Anthropic price as the anchor and adjusts it for differences between labs: governance control (trust or nonprofit), liquidation preferences, capped-profit history, compute-partner rights, and the thin-float premium. A new entrant would win by being the defensible, documented methodology that auditors accept for OpenAI, xAI and similar holdings during the next two or three quarterly marks. That probably means partnering with or selling to the valuation firms rather than going straight to LPs. It also depends on data access, because the investor-rights agreements are not public. The window may be short, because Caplight or 73 Strings could add a public-anchor adjustment quickly once Anthropic trades.
- Caplight (MarketPrice / Caplight Data)
Sells estimated prices for a wide range of private companies, and press reports use it as the source for the secondary prices of Anthropic (about $1.2T) and OpenAI (about $908B). It is the closest existing product to the reconciliation idea, but it reports raw secondary marks. It doesn't adjust for structure or rights against a public anchor.
- Forge Global (Forge Price)
Publishes a daily indicative price for about 200 pre-IPO companies, including the frontier labs, based on secondary trades. Forge itself says the price may rest on very few inputs. There's no rights-adjusted reconciliation to a public comparable.
- 73 Strings
Valuation and portfolio-monitoring software for private capital, backed by a $55M Series B led by Goldman Sachs Alternatives. It could add an Anthropic-anchored comparables module to the workflows funds already run, which makes it a likely fast follower.
- FactSet SharkRepellent
The established database of charter, bylaw and takeover-defense provisions, used by institutional investors. It will cover Anthropic's charter once it's public. It is not built around trusts, PBC structures or the frontier labs that are still private.
- ISS (governance and dual-class analytics)
Already scores dual-class and new-issuer governance for funds and index users, and is known for opposing directors at newly public companies with entrenched structures. It is the default incumbent for the governance-tracker buyer, but its coverage is generic rather than lab-specific.
- Anthropic IPO trackers (Leverage Shares, tastytrade, GraniteShares)
Free retail-oriented pages that track S-1 status, timing and valuation. They give no structured data on trust powers, share classes, lockups or float.
- Klover.ai Anthropic IPO governance deep dive
A one-off research piece on Anthropic IPO governance and control. Some of its share-class claims aren't backed by other sources. It shows interest in the topic, but it isn't a maintained data product.
- AI Lab Watch
A one-person project that scored frontier labs on safety and governance practices and is reported to be dormant. It is aimed at safety, not investors, and has no cap-table or float data.
- LongtermWiki (Anthropic LTBT entry)
Community wiki entries on the Long-Term Benefit Trust and lab governance. These are unstructured and not investor-grade, but they show the raw material is being collected by hobbyists.
Checked 2026-10-20
Risks
- The IPO timing could slip into a weaker market
- Governance concerns may be overshadowed by revenue growth
- Most of this is context for public investors, not a new startup category
Strong tech readiness and team fit; open space is the weak point (5).
tech 9 · pull 7 · space 5 · fit 9 · build 7
Exposed
Companies that could benefit or be threatened. News and context, not investment advice.
Momentum of its main topic
Sources
- OpenAI seeks to accelerate fundraising ahead of Anthropic IPO, sources sayr/stocks
- Sources: OpenAI is racing to secure investor commitments before Anthropic's planned IPO in mid-November, aiming to raise ~$30B; the round may value it at $1.4T+ (Reuters)Techmeme
- Anthropic commits 5% of shares to new public benefit fund that could address AI job disruptionCNBC
- The obscure trust looming over Anthropic's IPOHacker News
- Anthropic is AI trade's last big hope for a major IPO this yearCNBC
- CoreWeave bonds near 15%, Goldman upgrades anyway, and is the MBA finished? | Market HangYahoo Finance
- Anthropic announces public benefit fund, committing 5% of sharesCNBC Television
- Anthropic has nearly stopped hiring junior staff because AI agents now work six times as many hoursThe Decoder
Status history
- 2026-10-20 Flagged as buildable