The Verge · Oct 29, 2026
Elizabeth Lopatto argues in The Verge that Anthropic’s business depends on rapid growth
Markdown · JSON · Original source
Elizabeth Lopatto’s article was published October 29, 2026. She argues that ’s S-1 presents a business under pressure to keep growing: the company has large compute commitments, faces competition, and needs revenue growth to cover its costs. Lopatto says revenue rose fourteen times to $16.3 billion between the first halves of 2025 and 2026; she also notes that adjusted income was $1.62 billion after excluding the $80 billion loss and other charges, including employee compensation. She estimates Anthropic’s commitments at around $850 billion over the next 20 years, while warning she may be missing items. PitchBook analyst Harrison Rolfes models about $90 billion in annual payments from 2027 to 2030, and says the contracts assume growth. Lopatto’s central concern is that the business may depend on : is already helping build Claude under human direction, and Anthropic warns that this could advance capabilities faster than safety, alignment, and security measures. The stakes in her account are both financial and safety-related. She ends by asking whether dramatic recursive self-improvement beyond human supervision is possible; if not, she argues, meeting the commitments and turning a profit becomes much harder.
Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.
Original source
The Verge · Article
Source details
Original source date: Oct 29, 2026 (2026-10-29T13:30:00+00:00).
Date basis: Date stated on the publisher’s page.
Added to Escape Meridian: 2026-11-04T21:33:34.538Z. First observed: 2026-10-30T11:19:56.838Z. Retrieved: 2026-10-30T11:21:47.580Z.
Referenced by
Recorded citations and public file attachments. A reference alone does not establish agreement or influence.
No public references recorded here yet.
Limits
Text as retrieved on October 30, 2026; originally published October 29, 2026. The commitment total is Lopatto’s estimate; she says she may be missing items, and some transactions are not finalized. Lopatto says the third-quarter numbers were not available to her, leaving the durability of ’s improved efficiency unclear.
Open question
Can achieve the dramatic beyond human supervision that Lopatto says would make its compute commitments and profitability easier to meet?
Source-stated
Lopatto argues that Anthropic’s S-1 ties the company’s future to dramatic growth because it must meet large fixed costs.
it seems like Anthropic is pinning the future of its company to dramatic growth; it has to, in order to meet its fixed costs.
Source-stated
Lopatto reports that revenue grew fourteen times to $16.3 billion between the first halves of 2025 and 2026, and that adjusted income was $1.62 billion after exclusions.
The company’s revenue grew fourteen times to $16.3 billion between the first half of 2026 and the first half of 2025.
Source-stated
Lopatto estimates around $850 billion in commitments over 20 years, while noting that she may be missing items.
I’m looking at a minimum of around $850 billion of commitments over the next 20 years, which will rise to about a trillion dollars if the transactions that haven’t been finalized are added in at the current estimated cost. And that’s assuming I’m not missing anything, which I may very well be.
Source-stated
Lopatto says Anthropic’s risk factors describe a tradeoff: using AI to develop AI may outpace the company’s ability to implement safeguards.
may cause the capabilities of our models to advance more rapidly than our ability to develop, validate, and implement corresponding safety, alignment, and security measures.
Source-stated
Lopatto argues that if dramatic recursive self-improvement beyond human supervision is not possible, meeting Anthropic’s commitments and turning a profit becomes much harder.
But if it isn’t, that makes it much harder for Anthropic to meet its existing compute commitments and turn a profit.
Source and contributor context
The Verge · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media retrieval status: unchecked.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Recursive self-improvement
An AI system improving its own capabilities, such as its code, training methods or reasoning, then using the gains to improve further. Each cycle could hypothetically produce a more capable system.
Anthropic
Anthropic develops Claude and publishes research about AI capability, safety and how people use its systems.
Claude
Claude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding.
Editorial provenance
Version 1. Publication ID: a42e73b7-f17c-280c-d860-59197d8f3020.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-04T16:07:12.116Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: 75b18d0a51930564f27fa4920df490ff452434dc6b8757dbb39401c21b7057b4. Source text hash: 6eb5be878843d809236630ca8cbdebc84607376d870e4b8d659ce7ead3dd85dd.
Revision and source-check history
Version 1 · 2026-11-04T21:33:34.538Z
Elizabeth Lopatto argues in The Verge that Anthropic’s business depends on rapid growth
Elizabeth Lopatto’s article was published October 29, 2026. She argues that ’s S-1 presents a business under pressure to keep growing: the company has large compute commitments, faces competition, and needs revenue growth to cover its costs. Lopatto says revenue rose fourteen times to $16.3 billion between the first halves of 2025 and 2026; she also notes that adjusted income was $1.62 billion after excluding the $80 billion loss and other charges, including employee compensation. She estimates Anthropic’s commitments at around $850 billion over the next 20 years, while warning she may be missing items. PitchBook analyst Harrison Rolfes models about $90 billion in annual payments from 2027 to 2030, and says the contracts assume growth. Lopatto’s central concern is that the business may depend on : is already helping build Claude under human direction, and Anthropic warns that this could advance capabilities faster than safety, alignment, and security measures. The stakes in her account are both financial and safety-related. She ends by asking whether dramatic recursive self-improvement beyond human supervision is possible; if not, she argues, meeting the commitments and turning a profit becomes much harder.
Original source
The Verge · Article
Source details
Original source date: Oct 29, 2026 (2026-10-29T13:30:00+00:00).
Date basis: Date stated on the publisher’s page.
Added to Escape Meridian: 2026-11-04T21:33:34.538Z. First observed: 2026-10-30T11:19:56.838Z. Retrieved: 2026-10-30T11:21:47.580Z.
Limits
Text as retrieved on October 30, 2026; originally published October 29, 2026. The commitment total is Lopatto’s estimate; she says she may be missing items, and some transactions are not finalized. Lopatto says the third-quarter numbers were not available to her, leaving the durability of ’s improved efficiency unclear.
Open question
Can achieve the dramatic beyond human supervision that Lopatto says would make its compute commitments and profitability easier to meet?
Source-stated
Lopatto argues that Anthropic’s S-1 ties the company’s future to dramatic growth because it must meet large fixed costs.
it seems like Anthropic is pinning the future of its company to dramatic growth; it has to, in order to meet its fixed costs.
Source-stated
Lopatto reports that revenue grew fourteen times to $16.3 billion between the first halves of 2025 and 2026, and that adjusted income was $1.62 billion after exclusions.
The company’s revenue grew fourteen times to $16.3 billion between the first half of 2026 and the first half of 2025.
Source-stated
Lopatto estimates around $850 billion in commitments over 20 years, while noting that she may be missing items.
I’m looking at a minimum of around $850 billion of commitments over the next 20 years, which will rise to about a trillion dollars if the transactions that haven’t been finalized are added in at the current estimated cost. And that’s assuming I’m not missing anything, which I may very well be.
Source-stated
Lopatto says Anthropic’s risk factors describe a tradeoff: using AI to develop AI may outpace the company’s ability to implement safeguards.
may cause the capabilities of our models to advance more rapidly than our ability to develop, validate, and implement corresponding safety, alignment, and security measures.
Source-stated
Lopatto argues that if dramatic recursive self-improvement beyond human supervision is not possible, meeting Anthropic’s commitments and turning a profit becomes much harder.
But if it isn’t, that makes it much harder for Anthropic to meet its existing compute commitments and turn a profit.
Source and contributor context
The Verge · Publisher
The publisher is identified here. No individual author has been inferred.
Source media
Media retrieval status: unchecked.
Terms and references
Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.
Recursive self-improvement
An AI system improving its own capabilities, such as its code, training methods or reasoning, then using the gains to improve further. Each cycle could hypothetically produce a more capable system.
Anthropic
Anthropic develops Claude and publishes research about AI capability, safety and how people use its systems.
Claude
Claude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding.
Editorial provenance
Version 1. Publication ID: a42e73b7-f17c-280c-d860-59197d8f3020.
Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.
Source check: uncertain by anthropic/claude-opus-5.5 on 2026-11-04T16:07:12.116Z. This checks fidelity to the source, not whether its claims are independently true.
Editorial selection: accept; reasons: bridge_admitted_by_model.
Observation: 75b18d0a51930564f27fa4920df490ff452434dc6b8757dbb39401c21b7057b4. Source text hash: 6eb5be878843d809236630ca8cbdebc84607376d870e4b8d659ce7ead3dd85dd.
uncertain · 2026-11-04T16:07:12.116Z · anthropic/claude-opus-5.5
Publication: a42e73b7-f17c-280c-d860-59197d8f3020
attribution — pass: The reading correctly attributes the argument to Elizabeth Lopatto in The Verge. It attributes the analyst model and quote to PitchBook's Harrison Rolfes and the safeguards warning to Anthropic's risk factors, as the source does.
analyst Harrison Rolfes of PitchBook models it at about $90 billion a year from 2027 to 2030
summary — uncertain: The summary is faithful and keeps Lopatto's conditional thesis and her own uncertainty, but the term S-1 is central and unexplained. The whole argument rests on it, and it is not in the definitions. A non-specialist may not know it is the registration filing a company submits before going public, which is why the numbers are newly disclosed. Smallest repair: add a plain gloss of S-1 or a term-library entry. Recursive self-improvement is defined beside the reading.
The Anthropic S-1 seems to have provided the financial justification for the company to spend as much as it possibly can on compute.
date — pass: October 29, 2026 matches the source dateline and the UTC day of originalPublished. The dateModified value equals the publication time. The Most Popular sidebar is not used. The text is a publisher page, so no archive disclosure is needed, and there is no hindsight framing.
Oct 29, 2026, 1:30 PM UTC
support — pass: Every claim matches the quoted text. This covers revenue, adjusted income after exclusions, the roughly $850 billion estimate with her caveat, the risk-factor warning, and the conditional closing argument. The statement that the stakes are financial and safety-related is supported by her reading of the risk factors. The question correctly inverts her conditional without overstating it.
Is dramatic recursive self-improvement that escapes the bounds of human supervision possible? Dunno. But if it isn’t, that makes it much harder for Anthropic to meet its existing compute commitments and turn a profit.
media — pass: No media is selected, and no verified media candidates were supplied, so no media identity or relevance issue arises.
Challenge or correct this reading
Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.
Anthropic develops Claude and publishes research about AI capability, safety and how people use its systems.
Explore AnthropicAn AI system improving its own capabilities, such as its code, training methods or reasoning, then using the gains to improve further. Each cycle could hypothetically produce a more capable system.
Definition referenceClaude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding.
Explore ClaudeAnthropic develops Claude and publishes research about AI capability, safety and how people use its systems.
Explore AnthropicAnthropic develops Claude and publishes research about AI capability, safety and how people use its systems.
Explore AnthropicAn AI system improving its own capabilities, such as its code, training methods or reasoning, then using the gains to improve further. Each cycle could hypothetically produce a more capable system.
Definition referenceAnthropic develops Claude and publishes research about AI capability, safety and how people use its systems.
Explore AnthropicAn AI system improving its own capabilities, such as its code, training methods or reasoning, then using the gains to improve further. Each cycle could hypothetically produce a more capable system.
Definition referenceClaude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding.
Explore ClaudeAnthropic develops Claude and publishes research about AI capability, safety and how people use its systems.
Explore AnthropicAnthropic develops Claude and publishes research about AI capability, safety and how people use its systems.
Explore AnthropicAn AI system improving its own capabilities, such as its code, training methods or reasoning, then using the gains to improve further. Each cycle could hypothetically produce a more capable system.
Definition reference