Tuesday, 27 October 2026 Updated hourly from 226 outlets
Pricing Sign in

7 sourcesFirst reported yesterday at 13:05, updated 19 h ago

Anthropic’s IPO filing shows explosive growth, risks and staggering costs

Photo: South China Morning Post

Neutral summary

from 5 outlets' reporting

Anthropic has filed IPO papers for a Nasdaq listing under the ticker "ANT", expected before Thanksgiving, which could be among the largest IPOs on record. The prospectus reports a first-half 2026 net loss of $80.7 billion, largely from a rise in the fair value of convertible bonds, and revenue of about $16.3 billion. Revenue grew 1,368 percent in the three months ended June 30 from a year earlier, though the company warned growth could decline. The filing details risks including misuse, security threats and AI behaviours such as "resembling blackmail". Anthropic said more than half of Fortune 500 companies are customers. Some investors value the company at about $1.5 trillion, down from roughly $2 trillion a week earlier, Reuters reported. CEO Dario Amodei has called for slowing advanced AI development. Goldman Sachs and Morgan Stanley are lead underwriters; both declined to comment.

Where accounts differ. Al Jazeera, Channel NewsAsia and The Indian Express report the $80.7 billion loss figure, while bdnews24 and South China Morning Post give no loss or revenue figures, and only Channel NewsAsia and The Indian Express name the underwriters.

The Indian ExpressSouth China Morning PostChannel NewsAsia (CNA)bdnews24.comAl Jazeera English

Written by AI from these outlets' articles: facts they report, claims attributed, numbers checked against the sources. It may still contain errors; read the originals.

Article analysis

8out of 10

Band DEditorialising in its own voice

Reports the filing factually with attributed figures and risks, but dresses headline and narration in recurring intensifiers such as explosive, staggering, breakneck and red-hot.

Seven of the seven outlets that covered this event headlined it neutrally.

The paper's own narration evaluates, which places it in the top band.

1 Straight reporting10 Campaigning

The same event, one framing

The earliest headline of each kind. Marked words are judgements in the headline itself: stated as fact or in quotation marks.

Neutral 7 of 7

Anthropic's IPO filing shows explosive growth, risks and staggering costs

Channel NewsAsia (CNA)

How the 8 was worked out

The model answers seven fixed questions from 0 to 3 and cites evidence for each. The score is calculated from the answers in code, so any disagreement points to one question.

  1. G1
    Reporter's own voiceRecurring evaluative language in the reporter's voice. Evidence 2 3 4 5
    2
  2. G2
    HeadlineCompresses or intensifies, but the claim stays attributed. Evidence 1
    1
  3. G3
    Other side missingEvery party involved responds substantively.
    0
  4. G4
    Weight of the unanswered allegationNo allegation against an identifiable party.
    0
  5. G5
    Loaded quotesQuotes are informational.
    0
  6. G6
    Unattributed claimsEvery disputable claim is attributed.
    0
  7. G7
    Missing contextNecessary background is present.
    0

Evidence

1Emotional sensationalismHeadline voice

explosive growth, risks and staggering costs

Intensifiers 'explosive' and 'staggering' in the headline go beyond descriptive reporting of a filing.

2Emotional sensationalismReporter voice

grew a staggering 1,368 percent

Adds an evaluative intensifier to a figure that already conveys the scale.

3Word choiceReporter voice

Big Tech’s breakneck spending

Charged descriptor in narration where a neutral term would carry the same information.

4Word choiceReporter voice

lofty valuations

Evaluative adjective about valuations used in the paper's narration.

5Word choiceReporter voice

Red-hot growth

Evaluative subheading in the paper's voice.

Show the calculation
S = G2 + … + G7 = 1 + 0 + 0 + 0 + 0 + 0 = 1
G1 = 2, at least 2: band D
Band D: set by G1 and S. Score 8

Every headline, in order

About this method

Method. Bias types follow the Table of Media Bias Elements and the Media Bias Taxonomy (both CC BY 4.0). The impartiality rules follow the BBC Editorial Guidelines and the Ofcom Broadcasting Code. This analysis was made by claude-opus-5-5 on rubric 1.0.0. The score is calculated from the seven answers, never generated.

What this is not. No outlet gets a standing bias rating, and there is no left/right axis, because that axis does not describe Bangladeshi politics. Each analysis covers one article about one event. Members' analytics count how each outlet's headlines were labelled over a chosen period, with the number of headlines behind every share, and also show the average score of an outlet's scored articles: that average describes those articles (mostly each story's lead), not the outlet, and outlets are never ranked by it or by any share.

Known limits. Headlines are grouped into stories automatically, so an unrelated headline sometimes slips in; it is shown faded rather than hidden. Where outlets carry the same wire copy, the comparison has little to show. Human annotators agree on bias in Bangla news only moderately (κ = 0.574 at best), which limits what any automated score can claim.

The full method