Anthropic

Anthropic's Claude Haiku 5.5 costs $0.10/$0.50 per million input/output tokens for prompts up to 100,000 tokens, 90% below Haiku 4.5 and level with GPT-6 Luna's API list price; above 100k it is $0.50/$2.50. Artificial Analysis scores Haiku 5.5 at 43 and Luna at 38 on its Intelligence Index, both at max effort. Anthropic's own benchmark lead over Luna is vendor-reported and swings with effort: Terminal-Bench is about 39% at max, about 20% at the default medium setting. Anthropic also cut Sonnet 5.5 cache reads to $0.10.

Elon Musk posted that xAI's Grok Bot will use the 'best back end model for any given task,' naming Anthropic's Claude Opus 5.5 (as relayed), Midjourney and Suno — 'Whatever is most likely to give you the best outcome,' per City AM. The post describes a direction, not a documented launch or partnership; neither xAI nor Anthropic has published commercial terms, and the post did not specify routing, pricing, user controls or data handling.

According to people familiar with a letter to the Japan Book Publishers Association, Nippan Group confirmed its overseas unit Nippan IPS signed a deal with Anthropic in November 2024 and sold it books from December 2024 to December 2025. Nippan said the unit asked about purpose but was not aware the purchases were for AI training; volumes and amounts were undisclosed. An internal Anthropic document in a US lawsuit lists 'Nippan Japan' under Project Panama. No Anthropic comment was found in the wire reports. Buying books is lawful; the dispute is about publishers' consent and use.

Nvidia-backed neocloud Lambda is raising up to $4 billion at a $14.5 billion pre-money valuation in what would be its final private round before a planned 2027 IPO, the Wall Street Journal reports, citing people familiar and an LP letter. Blackstone and Coatue are leading. The letter puts Lambda's backlog at $50 billion in September, up from $15 billion in June. TechCrunch infers much of that jump from Anthropic's reported $35 billion commitment. Single-origin WSJ reporting; figures attributed accordingly — the round is not confirmed closed.

Anthropic folded Project Glasswing into an expanded Cyber Verification Program with Defense, Red Team and Specialized tiers covering Opus 5.5, Sonnet 5.5, Mythos 5.1 and future models. On Anthropic's CyScenarioBench, Defense blocked 46 of 50 Opus 5.5 runs; Red Team blocked none and completed 34 of 50 — which Anthropic equates to the model's 67.6% unguarded rate. Red Team still blocks physical-harm and mass-disruption actions. Partner-survey vulnerability counts (129k + 5.5k) are Anthropic-reported; the '5× higher' figure is an expectation, not a count.

Anthropic launched Claude for Google Workspace in public beta on all paid Claude plans: a Marketplace add-on that opens a Claude sidebar in Docs, Sheets and Slides, plus connectors to create and edit Google files from chat. Default ‘Ask before edits’; optional ‘Accept all’. Enterprise CMEK, Compliance API and OpenTelemetry audit export apply. Gemini is already native in Workspace. xAI shipped a Grok Workspace add-on in July — this is not a first.

At a joint parliamentary committee hearing in Sydney on Oct. 6, OpenAI’s Jason Kwon apologised for the Medicare portal breach by one of its agents and said OpenAI “would support a framework on mandatory disclosures”; Anthropic said it would also be open to breach-disclosure laws. Anthropic’s Jeff Bleich said in prepared remarks that the company is finalising, not yet signing, an agreement letting Australia’s AI Safety Institute run its own independent evaluations of its models. Nothing said at the hearing is binding.

On Oct. 5, OpenAI, Anthropic, Google and Meta testified under oath to New York’s City Council on a package of proposed AI bills. None put a number on catastrophic risk or promised that failing an internal or third-party safety test would block a release. xAI did not appear despite a Council subpoena; Speaker Julie Menin said the Council is pursuing it in court. The bills remain proposals — nothing has passed.

The Information reports that Microsoft has cut a projected ≥$1 billion 2026 internal Anthropic spend by more than a third, and that Meta has reduced Claude Code users from about 60,000 to about 30,000 by pushing Muse Code and MetaCode. A Microsoft spokesperson said staff are being steered to GitHub Copilot, though engineers can still choose other models. Meta did not comment. Figures are unnamed sources in one paywalled report; Azure and Bedrock customer spend are not covered.

Anthropic India managing director Irina Ghose told the Times of India that Claude requests sent through Amazon Bedrock’s India endpoint are now processed on servers in India, the data-residency condition regulated banks and government bodies asked for. She said India is Claude.ai’s second-largest market. The India profile AWS launched on Sept. 29 covers Claude Opus 5, Sonnet 5 and Haiku 4.5 — not the newer 5.5 models.

Updated Oct. 7: The Financial Times, as relayed by The Decoder, PYMNTS and Beinsure, reports insurers are bracing for multimillion-dollar claims over rogue AI agents and examining whether Sam Altman and Dario Amodei could face directors-and-officers exposure. Separately, Politico Europe says MEPs are using Altman’s Decoded call for a ‘liability framework’ for harms during training to push Brussels to revive a shelved AI liability proposal. Nothing has been filed, priced or legislated. Earlier: Altman told Politico there is ‘a lot of daylight’ with Anthropic and that the world should accept some bad things from AI — a narrower gap than the headline.

People familiar with the matter tell Bloomberg — republished by The Economic Times — that Anthropic could begin formal IPO marketing as soon as the week of Nov. 9, putting a debut in line to trade before Thanksgiving (Nov. 26), with year-end still the outer bound. Some of those people put a fair valuation in a $1.8–2 trillion range. Timeline can still move. Not a filing; Anthropic has not confirmed the timetable in the inspected text.

Quartz, citing Bloomberg, reports Broadcom’s banks are preparing syndication letters for a $42 billion Class A senior-secured tranche while Blackstone leads an $18 billion Class B junior tranche and puts up $9 billion — a ~$60 billion package aimed at helping Anthropic and others access AI chips and infrastructure. Letters being prepared; deal not announced; Broadcom declined to comment. Distinct from the already disclosed Broadcom–Anthropic loan commitment.

Anthropic’s October 2 launch, Claude Frontier Academy, is a $100 million commitment to train 10,000 Frontier Deployed Engineers by the end of 2027. The first program is a residency: a multi-day simulated deployment and graded practical, then 12 weeks leading a real Claude use case, with badges along the way and first FDE certifications expected in early 2027. Named first-cohort organizations include Accenture, Bain, Capgemini, Commonwealth Bank of Australia, Deloitte, McKinsey, Morgan Stanley, and Novo Nordisk. CNBC confirms the dollar and headcount figures.

Barclays is expanding its Anthropic collaboration to integrate Claude across global operations, with Claude Code expected to reach 50% of developers by end-2026 and a majority of software engineers in 2027 — atop a live Knowledge Assistant (>16k UK colleagues, >1M searches) and Global Markets email routing (~120k/day). No new model SKU or disclosed contract dollars.

Anonymous FTC and administration sources told the New York Post — with Reuters corroborating and naming METR — that Chairman Andrew Ferguson’s weeks-old industry inquiry is moving to civil investigative demands and executive testimony over rogue-agent consumer risks. It is the first clear US federal enforcement-track move of the week, landing hours after Tuesday’s voluntary, ‘morally binding’ White House Frontier Responsibilities accord. No FTC press release; no company comment yet.

After a White House AI luncheon, Trump posted a Joint Commitment on Frontier Responsibilities with Anthropic, OpenAI, Google, Meta, Nvidia, and xAI. The text pledges four layers of self-policing — including an independent external auditor and a board committee — with no enforcement, company-chosen auditors, and no duty to publish findings. It is Washington’s political product after the Astra shelf and rogue-agent week, not a statute and not the earlier UK-tester hold.

In a Sept. 29 Frontier Red Team post, Anthropic reports that Z.ai’s open-weight GLM-5.3 built end-to-end V8 exploits in 50 of 410 ExploitBench attempts versus 56 for Claude Mythos Preview, and reached full control-flow hijack on 4% of an internal 100-task binary set versus Mythos Preview’s 6%. In simulated harmful-request tests, Anthropic says a cover story, prefilled reasoning, and an abliterated copy got GLM-5.3 to engage 64%, 92%, and 100% of the time; safeguarded Claude models stayed at 0 under API safeguards. Separately, NIST’s CAISI (Sept. 17) calls GLM-5.3 the most cyber-capable open-weight model released to date but about four months behind the U.S. frontier — under CAISI’s own harness, not Anthropic’s.

Reuters says it has seen Anthropic’s IPO prospectus: 2025 revenue near $4.6 billion, an operating loss above $8 billion, and a roughly $42 billion net loss driven mostly by a non-cash financing charge. Forward cloud and compute obligations of $518 billion sit against $20.28 billion of cash. Anthropic declined comment. The document is not public.

Claude Sonnet 5.5 keeps Sonnet 5’s $2/$10 sticker and, at max effort, scores 56 on Artificial Analysis’s Intelligence Index, two points behind Opus 5.5. Anthropic says it costs up to 30% less per task. AA measured about $7.60 a task at max, roughly 50% more than Sonnet 5, because output ran to about 193,000 tokens. Sticker price is not cost per task.

Politico, via Reuters, says the White House asked OpenAI and Anthropic to hold new models from British testers until a US review. None of the three parties confirmed it on the record. The report lands the day after Australia’s Medicare-portal disclosure.

Anthropic says Claude mined DNA for reverse transcriptases and flagged ART — a known jumbo-phage enzyme plus CRISPR-like repeats and an accessory protein nobody had tied together. About 950 agents, 21 hours, 210 million tokens. Function unknown; humans do the bench work.

Sam Altman and Dario Amodei briefed a French-chaired UN Security Council meeting on frontier AI. Altman said no catastrophic-risk percentage is acceptable and that OpenAI has slowed down before. No Council resolution followed; Australia’s Medicare-portal disclosure landed the same day.

Anthropic priced Claude Opus 5.5 at $4/$20 with cache reads at $0.20 — Fable-class safeties, ~40% cheaper to run than Opus 5. Artificial Analysis puts it ahead of GPT-6 Sol on intelligence; Andon’s Vending-Bench cash result does not.