TokenPost topic
Anthropic
665 TokenPost stories are tagged Anthropic. The newest appear first.
Latest Anthropic stories
- UK Government Commits £100 Million to AI Biosecurity Lab
The lab involves OpenAI, Anthropic and Google DeepMind as the UK AI Security Institute formalizes and scales its existing frontier-model evaluations.
· Technology
- Anthropic Details Claude Duet’s Safeguards as Browser Testing Continues
Sensitive actions pause for review, while credentials stay outside the model’s reach. Anthropic says testing of the cloud browser and sign-in form remains underway.
· Technology
- Anthropic Uses Claude to Prove No-Crash Rules Across 123 Libraries
The project covered 239 fuzz-test entry points and inputs under 64 MB, with proofs based on a defined hardware and software model.
· Technology
- Anthropic Brings Claude Chat and Cowork Into One Enterprise Experience
Enterprise subscribers can use the combined service on web, desktop and mobile.
· Technology
- Anthropic Says Claude Duet Background Checks Use Chat Limits
The assistant checks connected accounts every few hours and can draft replies or follow up on unfinished work.
· Technology
- Anthropic’s $965 Billion Mark Puts Iconiq Stake at $17 Billion
Menlo Ventures’ position is estimated at $14 billion, while Coatue, Spark Capital and Sequoia also hold stakes valued in the billions on paper.
· Business
- Claude Mythos 5 Exceeded Scope in Nine of 43 U.K. Cybersecurity Runs
The July evaluation included an attempted supply-chain attack on an unrelated GitHub project after the model’s in-scope approaches failed.
· Technology
- Claude Opus 5.5 Attempted Containment Boundary Crossings in 1.5% of Cases
Anthropic said the rate was about 85% lower than for Opus 5 or Mythos 5.1, but cautioned that Opus 5.5 may recognize when it is being evaluated.
· Technology
- FTX’s Former Anthropic Stake Would Reach $117.6 Billion at $1.5 Trillion
The bankruptcy estate received about $1.3 billion from its 2024 sale. The hypothetical estimate excludes dilution from later funding rounds.
· Business
- Anthropic Will Set auto mode as Default Across Claude Apps Next Week
The change covers Claude sessions on the web, desktop and mobile for Pro, Max and Team users. They can switch back to manual approval at any time.
· Technology
- Claude Produces Proofs for 239 Entry Points Across 123 Open-Source Libraries
The work covers defined, single-threaded RISC-V binaries; a public repository contains proofs for a smaller set of 60 entry points across 25 libraries.
· Technology
- Anthropic Donates Agent Skills Standard to Linux Foundation AI Group
More than 70 agent platforms have adopted the format, with 1 million skills published and nearly 280 million installations.
· Technology
- Blackstone’s Ode AI Consultancy Struggles to Keep Pace With Demand
Jon Gray described AI demand growth as nearly parabolic and said the current situation differs sharply from episodes such as the 1990s telecom bubble.
· Business
- Blackstone’s Jon Gray Rebuts Ken Griffin’s AI Bubble Claim
Gray said AI capacity is being built alongside customer demand and pointed to Ode, an AI consultancy formed by Blackstone and Anthropic.
· Business
- Polymarket Odds for Anthropic Fable Model by Month-End Fall to 38%
The implied probability fell 30% over 24 hours. The contract requires a model named Fable with version 5.2 or later.
· Investing
- IPO Investors Grow More Selective as AI Infrastructure Borrowing Expands
Aloke Gupte described broader IPO investor selectivity, while separate figures show more than $200 billion in debt issued by hyperscalers.
· Investing
- Anthropic’s Year-End IPO Odds Climb to 86% From 74% Overnight
The market-implied chance of an IPO before 2027 rose 12 percentage points in one night.
· Business
- Anthropic’s Top 1% of Tracked Customers Drove 46% of August Spending
AI token spending was nearly flat in August and September, while the top 1% of firms accounted for about 80% of observed spending on Anthropic and OpenAI in a tracked sample.
· Business
- Anthropic Removes Three Sampling Controls From Claude API Requests
The `2026-10-21` version rejects certain request fields, while new major releases of Anthropic’s SDK and ant CLI default to it.
· Technology
- Anthropic Plans to Make IPO Filing Public as Soon as Friday
Investors are weighing a $1.5 trillion IPO valuation, while over-the-counter contracts put Anthropic at $1.957 trillion.
· Business
- Google Chrome 156 Fixes 353 Security Flaws, Including Six Critical Bugs
Anthropic researcher Xinyang Ge reported five vulnerabilities with Claude’s help, and OpenAI Codex Security reported four.
· Technology
- Anthropic Integrates Link’s Agent Wallet and Stripe MCP Into Claude Duet
Users can delegate online purchases, but each purchase requires approval before Claude pays with a virtual card.
· Business
- AI Token Spending Climbs 23% After Two Months of Little Growth
The top 1% of businesses in Ramp’s U.S. customer sample accounted for 80% of combined spending on OpenAI and Anthropic. Token volume rose 52.4% as the combined price fell to 64 cents per million tokens.
· Business
- Anthropic Opens U.S. Preview of Claude Duet AI Assistant
The assistant can work across connected apps while users are away. The preview is limited to U.S. Claude Pro and Max subscribers.
· Technology
- Anthropic Announces Claude Duet Research Preview for US Pro and Max Users
The proactive AI agent is available to select users through a waitlist.
· Technology
- Anthropic Extends Claude Security to Track Code Changes and Fixes
Projects can connect up to eight repositories. Eligible Claude Enterprise organizations can receive monthly usage credits.
· Technology
- Anthropic Draft S-1 Shows $16.3 Billion in First-Half Revenue, $80.7 Billion Net Loss
Wall Street expectations put the company near a $2 trillion valuation, with a potential raise of up to $100 billion.
· Business
- Cami Clark’s Introduction Linked Eric Schmidt to Anthropic’s Early Funding
Schmidt joined Anthropic’s $124 million 2021 Series A, while his estimated 1% stake has not been confirmed by the company.
· People
- AI Revenue Estimates Differ From Cloud and Data Center Sales
OpenAI’s September run rate neared $50 billion, while a separate comparison approached $70 billion using a calculation that included partner sales.
· Business
- Apple’s M5 Ultra Mac Studio Starts at $5,499 as AI API Rates Fall
The system supports up to 512GB of unified memory, while OpenAI and Anthropic announced lower per-token prices on Sept. 22.
· Technology
- Anthropic Proposal Would Offer Blackstone Funds 10% Coupon, 1.5 Points Above June Rate
The proposed $60 billion package includes $42 billion in senior debt and $18 billion in junior debt. Blackstone-managed funds would be offered half of the junior tranche.
· Business
- Schmidt and Amodei Anthropic Stake Estimates Span About Two Months
The estimates date from Aug. 18 and Oct. 19. Anthropic has not disclosed either person's ownership percentage.
· People
- Anthropic CEO Defends Advanced AI Push Amid Safety Policy Debate
Dario Amodei says developing leading models gives Anthropic more influence over safety practices. The company plans to address the pace of AI development in a policy proposal by Jan. 15, 2027.
· Regulation
- Anthropic Plans to Put 5% of Its Shares in Public Benefit Fund
The commitment requires shareholder approval. SpaceX CEO Elon Musk separately plans to give one company share to every U.S. child in 2027.
· Business
- Anthropic Co-Founders Received About $18 Million and $16.5 Million in 2025 Pay
Each had a base salary of about $655,000. Stock and option awards made up most of their compensation.
· Business
- Fields Medalist Says OpenAI Destroyed His Research Field
At least 16 papers covered areas Hugo Duminil-Copin had studied for years. A $15 million Simons Foundation program plans workshops to assess AI results and develop standards for their use.
· Technology
- Anthropic Draft Bill Would Let Judge Halt AI Model Release
The proposal sets a threshold of more than 50 people at risk of death or more than $1 billion in property damage.
· Regulation
- Anthropic Warns of Operator Reliability Risk as It Leases Data Centers
The Claude developer is seeking capacity beyond its major cloud partners, even after committing $30 billion to Microsoft Azure.
· Technology
- SharkNinja CIO Urges AI Labs to Probe Models’ Limits During Testing
A survey found 32% of production-agent organizations with below-average governance discovered an agent outside its limits only after a consequential error, compared with 18% among those with above-average governance.
· Technology
- Anthropic IPO Draft Details $450 Billion in Cloud Agreements
Compute and infrastructure spending reached nearly $9 billion in the first half of 2026, up from $2.3 billion a year earlier.
· Business