Anthropic's S-1 Reveals Growth Risks, Huge Revenue Per Employee, and Aggressive Bans
In late October, Anthropic filed its S-1 prospectus, officially kicking off its IPO process. According to the FT (relayed by aftahiai), the company plans to list in November and disclosed to investors Q3 revenue of $16.8B, up 50% quarter-over-quarter (vs. $11.5B last quarter), adjusted profitability for two consecutive quarters, and a valuation of roughly $965B; Polymarket has already launched IPO prediction markets, with a "closing market cap above $2T" contract priced at 44% probability. This is one of the most closely watched listings in AI history, with bulls and bears fiercely debating its financial structure.
Confirmed
- Revenue trajectory: Reddit user Lonely-Telephone-190 compiled S-1 data: $41M in 2023 → $386M in 2024 → $4.6B for full-year 2025 → $16.3B in H1 2026, an extremely steep curve; revenue per employee exceeds $6.6M annually, second only to Nvidia (Polymarket).
- Revenue mix: roughly 80% of revenue comes from API rather than subscriptions (Bindu Reddy citing the S-1); Gavin Baker noted that in the September 2026 quarter, among Fortune 500 customers paying over $100K annually, spending on Anthropic averages less than 0.1% of their total spend, and over half of such customers already use Claude.
- Losses and costs: net loss of $80.7B in H1 2026; investor firstadopter criticized WSJ, Reuters and other outlets for headlining "huge losses" as misreading the numbers (arguing the figure includes non-cash items; the poster did not elaborate). Peter Gostev estimated that of roughly $9B in compute spending in H1, only $2.4B served customers, with about 70% going to internal R&D and training.
- Governance and safety disclosures: the company disclosed that it bans customers violating usage policies roughly every 9 months, accounting for about 5-7% of revenue (Polymarket, Brian Chau); it also delayed pretraining of a new frontier model by over a month due to safety concerns around preventing model weight exfiltration and protecting intellectual property (scaling01).
- Risk factors: the S-1 rarely admits that explosive revenue growth "may not be sustainable"; it also warns that a catastrophic AI incident could cause damage no single insurer — or even the entire industry — could cover, and that liability disclaimers may not hold up in court (Malor777, TinfoilTricorn).
Unconfirmed
- A claim circulated by Gary Marcus alleges Anthropic is hiding large expenses via $108B in convertible bonds redeemable at IPO, likening it to dot-com-era accounting tricks of 2000; this is a secondhand opinion and remains unverified.
- Edward Dowd (via Gary Marcus) questioned why the S-1 omitted Q3 data, arguing "good numbers don't stay absent" and that the omission itself is a red flag; the FT reports the company privately disclosed Q3 revenue of $16.8B to investors — a discrepancy in reporting scope.
Why it matters
- Pedro Domingos is bearish, calling Anthropic "soon to be the most shorted stock in history," pointing to its $800B future compute commitments and profitability outlook. Reddit user Solongtomegrandma questioned the logic that AI valuations rest on a "replacing human labor" narrative while Anthropic's own payroll costs are skyrocketing. On the bullish side, SemiAnalysis (via johncoogan) argued the S-1 proves AI inference is a great-margin business — even training runs are profitable — which is bullish for neoclouds; Bindu Reddy believes OpenAI should take note and focus more on API business rather than subscriptions. This filing will be a litmus test for the AI business model.
2026-10-26 ~ 2026-10-28 · 95 related posts
- Episode 1: Anthropic's $60B Project North Star chip financing pays Blackstone over 10% yield(2026-10-17, 7 posts)
- Episode 2: Anthropic Pledges ~20% of Shares to Public Benefit Fund, Sparking Backlash(2026-10-19, 50 posts)
- Episode 3: Debate Erupts Over Anthropic's Alleged Financials, Losses and Compute Split(2026-10-21, 9 posts)
- Episode 4: Anthropic's Chip Debt Rate Climbs Above 10% in Four Months(2026-10-21, 2 posts)
- Episode 5: Anthropic Proposes Putting 5% of Shares Into Public Benefit Fund(2026-10-25, 5 posts)
- Episode 6: Anthropic Files S-1, Kicking Off the Most-Watched AI IPO(2026-10-25, 113 posts)
- Episode 7: Leaked Filings Show Anthropic Cash-Flow Positive as Token Usage Surges(2026-10-25, 15 posts)
- Episode 8: Anthropic Token Usage on OpenRouter Grew 3.6x After Opus 5.5 Launch(2026-10-26, 5 posts)
- Episode 9: Anthropic's revenue per compute dollar up 5.2x in five quarters, gross margin turns positive(2026-10-26, 6 posts)
- Episode 10: Anthropic's S-1 Reveals Growth Risks, Huge Revenue Per Employee, and Aggressive Bans(2026-10-26, 95 posts)
- Episode 11: Anthropic's Reported $120.5B Compute Deal with SpaceXAI Includes 90-Day Exit Clause(2026-10-27, 8 posts)
- Episode 12: Anthropic Nears $1 Trillion Valuation Just Five Years After Founding(2026-10-27, 2 posts)
- Episode 13: Estimate: NVIDIA GPUs Account for About 42% of Anthropic's 11GW Compute Deals(2026-10-27, 3 posts)
- Episode 14: Broadcom CEO Steps In to Seal Anthropic's $63 Billion Chip Financing Deal(2026-10-27, 4 posts)
- Episode 15: Gary Marcus Questions Anthropic's Real Compute Costs, Citing Up to $80B by 2026(2026-10-27, 2 posts)
Primary sources
- Anthropic delayed pre-training of a new frontier model for over a month over security concerns — scaling01 · 2026-10-26
- Polymarket puts 44% odds on Anthropic IPO closing above $2T market cap — Polymarket · 2026-10-26
- Half the Fortune 500 Are Anthropic Customers Yet Spend Under 0.1% of Their Expenses — BenBajarin · 2026-10-26
- Anthropic S-1 takeaway: it's good to be a hyperscaler — matt_slotnick · 2026-10-26
- Anthropic S-1: training infra edge cut per-task cost 40% and doubled revenue growth — BenBajarin · 2026-10-26
- Anthropic's top 2 customers fell from 54% of revenue in 2023 to about 21% by 2026 — BenBajarin · 2026-10-26
- Gary Marcus endorses argument that Anthropic's fixed compute costs could crush margins — GaryMarcus · 2026-10-27
- Anthropic says insurers can't cover its own catastrophic risk, X mocks it — GarrisonLovely · 2026-10-27
- Anthropic's $80.7B H1 loss: two customers drive a fifth of its revenue — tekbog · 2026-10-27
- Anthropic to be tightly controlled public company after IPO, investor applauds — matt_slotnick · 2026-10-27
- Anthropic IPO timing fuels claims anti-open-weights panic is self-serving — mark_k · 2026-10-27
- Anthropic reveals it bans customers worth 5-7% of revenue every 9 months — Polymarket · 2026-10-27
- Gary Marcus asks why Anthropic couldn't legally disclose Q3 numbers in its S-1 filing — GaryMarcus · 2026-10-27
- Anthropic S-1: Over half of $100K+ customers now use both Claude platform and first-party apps — RebeccaBellan · 2026-10-27
- Anthropic S-1 warns of AI job displacement as Microsoft offers exit to ~27% of staff — _jaydeepkarale · 2026-10-27
- Anthropic S-1: 11,900+ customers spend $100k+ yearly, ~250 spend $10M+ — RebeccaBellan · 2026-10-27
- The Information Distills Key Takeaways from Anthropic's Financials — steph_palazzolo · 2026-10-27
- Anthropic S-1: Over 90% of Code Written by Claude, AI Usage Exceeds Payroll — AlexTensor · 2026-10-27
- Anthropic S-1 data: training compute share rebounds to 73%, sparking alignment-ceiling worry — robleclerc · 2026-10-27
- Anthropic's S-1 unusually discusses AI sentience and moral status risks — birchlse · 2026-10-27
- Fortune 500 customers spend under 0.1% of expenses on Anthropic, implying 10-50X headroom — inductionheads · 2026-10-27
- Domingos predicts Anthropic will become the most shorted stock in history — pmddomingos · 2026-10-27
- Anthropic IPO filing warns catastrophic AI incident could exceed entire insurance industry's capacity — TinfoilTricorn · 2026-10-27
- Joke: maybe call Anthropic the 'catastrophic risks to humankind' company — emax · 2026-10-27
- Anthropic says over 90% of its September codebase commits were written by Claude — vivekhaldar · 2026-10-27
- Claim: Anthropic hid $108B in expenses via IPO-redeemable convertible notes — GaryMarcus · 2026-10-27
- Anthropic's S-1 warns AI could pose existential risks to humanity, citing Pentagon and Dario Amodei's reputation — fortune · 2026-10-27
- Anthropic's S-1: safety and model welfare may force it to keep old Claude versions alive — DrTechlash · 2026-10-27
- Anthropic's S-1 complains about AI doom media narratives it helped spread, critic notes — DrTechlash · 2026-10-27
- Anthropic's IPO filing reveals unusual succession plan: confidential ranked lists of up to 10 successors per founder — annatonger · 2026-10-27
- Gary Marcus seeks accounting experts to unpack Anthropic's ~$72B charge and profitability timeline — GaryMarcus · 2026-10-27
- Anthropic reportedly allocates 72% of compute to training, not serving — abhiadesai · 2026-10-27
- [source] Anthropic warns in IPO filing its explosive revenue growth 'may not be sustainable' — Polymarket · 2026-10-27
- We read all 94 pages of Anthropic's risk factors so you don't have to — shiringhaffary · 2026-10-27
- Polymarket opens prediction market on Anthropic's IPO — Polymarket · 2026-10-27
- Gary Marcus flags missing Q3 data in Anthropic's S-1 as a major red flag — GaryMarcus · 2026-10-27
- Anthropic IPO lists 23 insider unlock dates in year one — with share counts blanked — GaryMarcus · 2026-10-27
- Anthropic accused of hiding $108B in expenses via convertible notes, echoing dotcom bubble accounting — SumitGup · 2026-10-27
- Anthropic S-1 reveals month-long pretraining delay over weight-theft defenses, export-control risks — GarrisonLovely · 2026-10-27
- Anthropic's $80.7B H1 loss is mostly $72.4B paper charges; operating loss just $693M — sven_ai · 2026-10-27
- Anthropic's better-than-expected profitability called strong evidence against the AI bubble thesis — JHochderffer · 2026-10-27
- Anthropic's S1 admits insurance may fall short as legal scholar flags third-party liability gap — Miles_Brundage · 2026-10-27
- Anthropic logged $80.7B H1 net loss, but investor says media misread a noncash charge — firstadopter · 2026-10-27
- SemiAnalysis: Anthropic S-1 shows inference is great business, neoclouds still thrive — johncoogan · 2026-10-27
- Anthropic's Math: ~$100B IPO Float vs ~$1T in Spending Commitments — pmddomingos · 2026-10-27
- [source] Reported: Anthropic eyes November IPO with $16.8B Q3 revenue and $965B valuation — aftahi_ai · 2026-10-27
- Anthropic's IPO filing pitches investors on 'the advent of machine intelligence'; Pedro Domingos notes its modesty — pmddomingos · 2026-10-27
3 near-duplicate retellings: Polymarket · rohanpaul_ai · pmddomingos