Anthropic news
Summary
262 news articles and 4 research items in this section, back to 24 November 2025. Browse the news ↓ Research ↓
52 stories from 26 publishers in the last 24 hours, 119 from 45 this week; 4 stories were covered by more than one newsroom.
Anthropic’s IPO puts AI’s circular financing boom to the test
Anthropic faces enormous compute commitments as it goes public, but its biggest infrastructure partners, including Amazon and Google, could continue profiting from AI even if Anthropic loses its lead.
Also at Sifted
Anthropic Prices IPO Shares At $83, Above Range
The offering promises to be the public market event of the season, with the Claude creator seeing investor demand for shares well in excess of its capital-raising goals.
Also at TechCrunch
Anthropic Cuts Disclosure Deadline to 30 Days For Serious Bugs
Anthropic will publish serious flaws found by its AI models 30 days after reporting them
Also at Help Net Security
Most mentioned in Anthropic news headlines, last 3 days: Anthropic 82IPO 52Debut 13Claude 10Record 9raises 8Stock 7Above 7Price 7billion 7
Most active this week: Tech in Asia (14), The Decoder (10), Nikkei Asia (9), The New Stack (7). Built automatically from headlines; the top items are the stories the most newsrooms covered in the last three days.
Analysis
Anthropic heads into its planned Nasdaq debut on 19 November under the ticker ANT amid intense investor demand, even as its CEO concedes deep uncertainty over the company's enormous compute commitments.
The dominant story of the past 48 hours is Anthropic's imminent initial public offering, reported as potentially the largest in history and valued at around $1.4 trillion, with trading due to begin on Nasdaq on 19 November 612. Demand appears strong in Asia, where investors are seeking indirect exposure and Chinese buyers face restrictions on participation 112, but the company's own chief executive has described buying compute as its 'deepest business dilemma' and acknowledged 'huge uncertainty' about next year 6. Alongside the financial narrative, the newly released Claude Fable 5.2 is drawing mixed early reviews on cost-effectiveness 3710, and a supply-chain compromise of the official Claude Code extension on Cursor's marketplace has been traced to a stolen access key 45. Together these items sketch a company at a moment of maximum visibility, with its commercial, technical and security standing all under close scrutiny.
Key developments
- IPO set for 19 November with heavy Asian demand. Anthropic is scheduled to begin trading on Nasdaq under the ticker ANT on 19 November, in an offering described as a $1.4 trillion IPO [12]. Chinese investors face limits on participating, while Asian investors more broadly are chasing indirect routes to exposure, suggesting demand well in excess of what direct allocations will satisfy [1][12]. The offering is reported to be intended largely to fund the company's compute build-out [6]. 1612
- Amodei flags compute as the central business risk. Days before the listing, CEO Dario Amodei said there is 'a huge uncertainty in what's going to happen next year' and called buying compute 'the deepest business dilemma' the company faces [6]. Anthropic's commitments are reported at nearly $901 billion, and the report notes it remains unclear whether future revenue will justify those costs [6]. The candour is notable given the timing and frames the IPO as a financing exercise for an unresolved question of scale. 6
- Former researcher cites roadshow slide in call for a pause. Jacob Coxon, formerly of OpenAI and Anthropic, says a slide from Anthropic's IPO roadshow shows the company's algorithms improving at an accelerating rate, and that Anthropic can already retrain its Mythos model with 'a fraction of the compute originally needed' [11]. He gives this as a reason for leaving and continues to call for a pause on frontier development so competitors can catch up [11]. This is a single former employee's interpretation of a roadshow document and should be treated as one contested reading of Anthropic's trajectory, though it sits awkwardly alongside Amodei's own emphasis on compute scarcity [6][11]. 611
- Claude Fable 5.2 draws mixed cost-performance verdicts. Anthropic released Claude Fable 5.2 on 9 November, saying it beats Claude Opus 5.5 in almost every capability, but one evaluation found the two models scoring the same on its tests at almost four times the cost for Fable [3]. Separate comparisons reported fewer refusals and only small savings against Fable 5.1 [7], and that a lower list price than GPT-6 Astra did not translate into a lower bill [10]. The model's distinctive image approach, writing Python code to place hundreds of thousands of simulated brushstrokes rather than computing pixels directly, is drawing attention as a novel technique [2]. 23710
- Trojanised Claude Code extension traced to stolen key. The compromise of Anthropic's official Claude Code extension on Cursor's extension marketplace resulted from the theft of an access key that Cursor had issued to Fireworks AI, exposed following a breach at Fireworks [5]. Security coverage described the incident as Claude Code being turned into a 'remote-controlled implant', placing it among the week's most significant security stories [4]. The episode underlines how third-party credentials in the AI tooling supply chain can expose even official distributions of widely used developer tools [4][5]. 45
- Opus 5.5 becomes default as developer unease grows. Anthropic has made Claude Opus 5.5 the default model on all paid plans and published guidance urging users to abandon older prompting habits such as 'think hard' and to start at Medium effort instead [8]. At the same time, founders building on Claude are reported to be concerned about Anthropic's expansion into enterprise software, raising the familiar platform question of whether the provider will compete with the businesses built on it [9]. 89
Themes
- Record valuation versus acknowledged uncertainty. The IPO is being marketed at an unprecedented scale and attracting strong demand, yet the company's leader is publicly describing its core cost base as an unresolved dilemma [1][6][12]. A former insider's claim that algorithmic gains are reducing compute needs adds a further, contested, layer to how investors should read the economics [11]. 161112
- Headline claims meeting independent scrutiny. Anthropic's own messaging on Fable 5.2 and Opus 5.5 emphasises capability gains and lower prices, but third-party testing repeatedly finds the real cost picture more complicated than list prices or benchmark claims suggest [3][7][8][10]. This gap between stated and observed value is likely to shape developer decisions, particularly given existing unease about Anthropic's enterprise ambitions [9]. 378910
- Security exposure through the ecosystem. The Claude Code incident was caused not by a flaw in Anthropic's own systems but by a credential exposed at a third party and then abused via a marketplace distribution channel [5]. As Anthropic's tools become embedded across developer workflows, its security reputation increasingly depends on partners it does not control [4][5]. 45
What to watch
- The 19 November Nasdaq debut and first-day trading of ANT will test whether the reported $1.4 trillion valuation holds against the uncertainties Amodei himself has described [6][12]. 612
- Any further disclosure from Cursor, Fireworks AI or Anthropic on the scope of the trojanised extension and affected users will determine whether the incident has lasting reputational or regulatory consequences [4][5]. 45
- Whether Anthropic responds to independent findings that Fable 5.2 costs far more than Opus 5.5 for similar results, through pricing or guidance changes, will indicate how it manages its model line-up post-IPO [3][8][10]. 3810
- Reaction to Coxon's claims about accelerating algorithmic efficiency, including any clarification of the roadshow slide, matters because it bears directly on the compute-cost thesis underpinning the offering [6][11]. 611
Stories this analysis is based on (12)
- Asian investors chase indirect access to Anthropic shares amid IPO fever Nikkei Asia
- Anthropic's Claude Fable 5.2 doesn't generate images, it paints them The Decoder
- Claude Fable 5.2 vs. Opus 5.5: Same scores at almost 4x the cost The New Stack
- Week in review: Secret stealer hits GitHub repos, Claude Code turned into remote-controlled implant Help Net Security
- Stolen Fireworks AI Key Led to Cursor Data Breach and Trojanized Claude Code Extension The Hacker News
- Anthropic CEO Amodei admits "there's a huge uncertainty in what's going to happen next year" days before a record IPO The Decoder
- Claude Fable 5.2 vs. Fable 5.1: Fewer refusals, with small savings on the login fix The New Stack
- Anthropic says outdated prompting habits are slowing down Claude Opus 5.5 The Decoder
- Anthropic wants you to build on Claude. There’s a catch. The New Stack
- Claude Fable 5.2 vs. GPT-6 Astra: A lower list price didn’t mean a lower bill The New Stack
- Anthropic can already retrain Mythos with "a fraction of the compute originally needed," says former researcher Coxon The Decoder
- Chinese investors face limits on Anthropic’s $1.4t IPO Tech in Asia
Written by Claude Fable 5.1 (Anthropic), based on 12 recent headlines and excerpts in this section, 3d ago. AI analysis can misread or miss things: the numbers link to the original reporting, which is the authority.
Research 4
Journal articles, academic blogs, think-tank reports and papers, kept apart from the news below.
Same Papers, Different Verdicts: Tracing Version Drift in LLM Peer Review
Venues and benchmarks read LLM review scores as measurements of a paper, while providers replace their models within months. We audit twenty frontier models from OpenAI, Google and Anthropic on ICLR 2027 submissions while they were under review, with tests registered before the decisions. Between an earlier and a later version of Claude Opus and of Gemini, the share of papers placed on the accept side falls by 52 and 53 percentage points on 183 and 60 randomly drawn…
Reliability of Claude Code as an Agentic AI Grading System for Handwritten University Exams
This study examines the reliability of Claude Code, an agentic deployment of the Claude Opus 4.6 large language model, as a grading system for handwritten university exams. Four research questions addressed agreement with expert human grading on total scores (RQ1), agreement across five question types from multiple choice to evaluative judgement (RQ2), consistency across two independent grading sessions one week apart (RQ3), and the accuracy of AI-generated feedback (RQ4)…
Human and planetary health implications of AI-generated diet plans: a comparative assessment using the EAT-Lancet framework
IntroductionArtificial intelligence (AI) is increasingly being used to support dietary planning, with growing interest in generating recommendations that are both nutritionally adequate and environmentally sustainable. This study compared meal plans generated by three AI platforms, OpenAI ChatGPT (GPT 5.5 Instant), Google Gemini (Gemini 3 Flash), and Anthropic Claude (Claude Opus 4.8), across different clinical scenarios using the energy adjusted EAT–Lancet Diet Index 2.0…
Aegis OS: A Transactional Multi-Tenant Operating System Runtime for Eliminating Authorization Fatigue in Autonomous Coding Agents
Autonomous coding agents such as Google Antigravity (agy), Anthropic Claude Code (claude), and OpenAI Codex CLI (codex) execute multi-step software engineering workflows on workstations and cloud hosts. Existing agent environments face a Security-Autonomy Dilemma: operators must choose between per-action confirmation prompts, which interrupt workflows on 45%-57% of tool calls and induce authorization fatigue, or unconfined execution (--yolo / --dangerously-skip-permissions)…
All news 262

Asia’s stock markets dive as Anthropic debut fails to allay AI bubble fears
Indexes in Japan, South Korea, Taiwan and Hong Kong fall amid tech sell-off.

US stocks fall as Anthropic IPO rally loses momentum
Anthropic’s market value briefly topped US$2 trillion during a closely watched test of investor appetite for AI stocks.

India’s Anthropic Question, ShareChat’s Turnaround & More
Can India Build Its Own Anthropic? Anthropic’s blockbuster listing has put India’s AI ambitions under a harsh spotlight. The IPO…

Meet the 8 women powering India’s deeptech future; Anthropic raises a record $81.3B in IPO
On Women’s Entrepreneurship Day, we look at women founders pushing the boundaries of Indian deeptech, building frontier technologies, and opening new possibilities for innovation and entrepreneurship.

OpenAI backer SoftBank sinks 16% after Anthropic debut
The decline highlighted SoftBank’s exposure to OpenAI, Anthropic’s rival.

Anthropic’s 7 founders cashed out nearly $2B in record-breaking IPO
At the upgraded $83/share IPO price, Anthropic's seven founders each pocketed $280 million, while the company nabbed nearly $105 billion.

Anthropic’s big Wall Street moment
Anthropic’s IPO puts the spotlight on the next big test for the AI industry, while potatoes put AI to the test on the farm.

Anthropic rival says companies can (mostly) skip the pricey models
On the day Anthropic went public, Fireworks AI’s co-founder made the case for open-source models as a cheaper alternative, and addressed the breaches that have cost the company some customers’ trust.

Japanese investors seek $9.4b in Anthropic IPO shares
They were ultimately allotted US$2.98 billion of shares, according to a filing.

How did Anthropic's stock market debut unfold?
The BBC's Samira Hussain breaks down the AI company's first day as a publicly traded company, as shares closed up 17%.

Policymakers watch Anthropic IPO with wary eye
Presented by Google Public Sector. {beacon} Technology Technology The Big Story Policymakers watch Anthropic IPO with wary eye Anthropic went public on Thursday, opening with a massive valuation of about $2 trillion. But policymakers were watching with a wary eye, as the AI industry faces a moment of upheaval and uncertainty. © Adobe Stock The AI...

Anthropic closes 17% higher in Nasdaq debut
AI company Anthropic closed at US$97.17 in its Nasdaq debut on November 19. Shares opened at US$111.5.

Protesters rally against Anthropic in New York and San Francisco after $2tn stock market debut
Anti-AI groups gather in front of Nasdaq exchange in Times Square and Anthropic’s headquarters in San Francisco As the artificial intelligence startup Anthropic debuted on the US stock market at an eye-popping $2tn valuation on Thursday, protesters from anti-AI groups rallied in front of the Nasdaq exchange at its New York City headquarters in Times Square and several were arrested. Similar demonstrations took place outside Anthropic’s San Francisco headquarters. “I think…


Anthropic becomes most valuable AI firm as shares soar in stock market debut
Anthropic is the first purely AI company to be publicly listed, creating a test for the entire industry.


Three arrested during Anthropic IPO protest outside Nasdaq
As the AI giant Anthropic made its record-breaking stock market debut in New York, police arrested demonstrators

Anthropic shares stay above IPO price in volatile debut
At 1:15 p.m. in New York, Anthropic’s stock traded at about US$94.61, nearly 14% above its IPO price of US$83.

Anthropic’s IPO could deliver $156 billion to charity
At a stock price of $95, the frontier lab's Public Benefit Foundation, its employee donations, and the company's match tot up to $156 billion.

Anthropic opens at $111.50 per share after largest IPO ever
Artificial intelligence giant Anthropic opened at $111.50 per share on Wednesday, after the company wrapped up the largest initial public offering (IPO) on record. Anthropic, which is trading under the ticket “ANT,” and certain stockholders sold 980 million shares of the company’s stock valued at $83 per share, the AI firm announced Wednesday. That brought...


Amazon’s $18B bet on Anthropic tops $200B, for a moment, in the Claude maker’s historic market debut
Anthropic went public in a record $81.3 billion IPO, and Amazon's $18 billion investment was worth about $216 billion on paper at the $111.50 opening price. Microsoft is buying $1 billion in stock, and Anthropic has committed to spend at least $47 billion on Azure.

Markets react to Anthropic’s IPO
Watch a live look at the New York Stock Exchange big board as Anthropic trading begins.
Anthropic throws open doors at $2T valuation as shares pop 34% in initial trading
Company still has to content with massive compute obligations and the threat of open models, neither of which Dario Amodei is worried about, he claims

Anthropic hits NASDAQ exchanges at $111, 33 percent above IPO price
Anthropic's IPO price was set at $83, representing more than $80 billion in new capital raised.

A “headless Slack”: Vercel’s new message board lets Claude Code and Codex agents work together
For AI agents to work together effectively, they need a way to communicate and divvy up tasks — much as

Anthropic starts trading in the US after blockbuster IPO
Anthropic’s IPO surpasses SpaceX, marking the largest initial public offering in US history at $81bn

Anthropic starts trading on Wall Street
Anthropic has officially gone public. It began trading on the Nasdaq under the ticker symbol "ANT" at $111.50 per share, up from its initial $83 per share price that put company's value at around $1.5 trillion. The listing comes nearly a month after Anthropic revealed its plans to go public. In its lengthy S-1 filing,

Anthropic CEO acknowledges AI has 'very real risks' shortly before IPO – video
Anthropic's CEO, Dario Amodei, issued a warning about the dangers of AI, saying it has 'very real' but 'avoidable' risks, shortly before the company rang the opening bell at Nasdaq Anthropic ‘to open up more than 20% higher at over $100 a share’ after blockbuster IPO – business live


