# Anthropic Watch Log News and developments about [[Anthropic]], newest first: funding and its planned stock listing, the legal fight over the federal ban, government deals, model releases and key hires. Each entry says what happened, why it matters and where the report came from. For the full picture, start with the [[Anthropic Dossier]]. > [!note] Entries currently run through August 27, 2026. ## Log ### 2026-08-27 — Anthropic wins outright: summary judgment, both orders vacated, permanent injunction, no stay **Category:** Government/Legal Judge **Rita F. Lin** (N.D. Cal.) granted Anthropic **summary judgment** in a **59-page order** on [[First Amendment]], [[Fifth Amendment]] due-process and [[Administrative Procedure Act]] grounds. She **vacated the supply-chain-risk designation**, **vacated [[Pete Hegseth]]'s order barring military contractors from any commercial business with Anthropic**, entered a **permanent injunction**, and **denied** the government's request for a seven-day administrative stay — noting it had operated under her March preliminary injunction for five months without identifying a single concrete harm. Lin found the measures rooted in a desire to make a public example of Anthropic for its "arrogance," not in evidence of any security threat: "The empty invocation of national security is not a blank check to punish and retaliate against government critics." Not a clean sweep. Lin **rejected** Anthropic's claim that [[Donald Trump]]'s directive exceeded presidential authority, and entered judgment **for** agencies that had taken no action. The holding governs *how* the executive acted, not whether it may direct agency AI procurement. **Why it matters:** the vacated **boycott order** is the piece with reach beyond the parties — it had barred defense contractors from dealing with Anthropic at all. [[Palantir]] is the largest incidental beneficiary: the [[Maven Smart System]] runs Claude via [[Palantir AIP]], and the order had pointed Palantir toward a costly model swap inside its flagship Pentagon program. See [[Anthropic Government Ban]] for the full case history. ⚠️ Appeal posture unreported. Procedural history is split — the **D.C. Circuit denied a stay of the designation in April** while N.D. Cal. has now vacated it. The [[Anthropic Model Export Directive]] (June 2026 Commerce "cyber weapons" classification) is a separate track and is untouched by this ruling. Source: [Judge rules the Pentagon's supply chain risk label for Anthropic unlawful (CNN, 2026-08-27)](https://www.cnn.com/2026/08/27/tech/anthropic-pentagon-supply-chain-risk-unlawful-hnk) · [Judge strikes down Hegseth's ban on Anthropic AI (Washington Times, 2026-08-27)](https://www.washingtontimes.com/news/2026/aug/27/judge-strikes-pete-hegseth-ban-anthropic-ai-calls-unconstitutional/) · [Judge blocks Pentagon blacklist of Anthropic AI (Axios, 2026-08-28)](https://www.axios.com/2026/08/28/judge-blocks-pentagon-anthropic-blacklist) ### 2026-06-27 — Mythos limited release approved **Category:** Government/Legal, Products Trump administration allowed Anthropic to release Mythos 5 to 100+ U.S. institutions (companies and government agencies). Fable remains unrestricted. Anthropic had earlier disabled access to both Mythos and Fable to comply with government orders suspending use by foreign nationals — including Anthropic's own employees. **Source:** CNN, CNBC, Semafor **Significance:** First concrete sign of rapprochement since the ban. Government is selectively restoring access while maintaining the FASCSA designation. ### 2026-06-19 — John Jumper hired from Google DeepMind **Category:** Talent Nobel laureate [[John Jumper]], co-creator of [[AlphaFold]] (2024 Chemistry Nobel), left [[Google DeepMind]] for Anthropic. Several AlphaFold collaborators reportedly following. Same week [[Noam Shazeer]] left DeepMind for [[OpenAI]]; twin departures erased $200B+ from [[Alphabet]] market cap. **Source:** CNBC, TechCrunch **Significance:** Highest-profile AI research hire of 2026. Signals Anthropic's push into AI-for-science beyond LLMs. ### 2026-06-09 — Fable 5 GA, Mythos 5 limited release **Category:** Products Claude Fable 5 became generally available. Claude Mythos 5 offered in limited availability to approved customers (Project Glasswing). Introduced the Mythos-class tier above Opus. **Source:** Anthropic newsroom **Significance:** New model tier architecture. Mythos restricted access creates a two-track product: public (Fable) and cleared (Mythos). ### 2026-06-01 — Confidential S-1 filed **Category:** Funding/IPO Anthropic confidentially submitted draft S-1 to the SEC for proposed IPO. Targeting Nasdaq listing as early as October 2026. Filed at $965B valuation. **Source:** Anthropic announcement, TechCrunch, NPR **Significance:** Three days after Series H closed. Revenue run-rate ~$47B annualized. Cleaner corporate structure than [[OpenAI]]'s S-1 (no nonprofit conversion baggage). ### 2026-05-28 — Series H: $65B at $965B valuation **Category:** Funding/IPO Largest AI funding round to date. Led by Altimeter Capital, Dragoneer, Greenoaks, Sequoia. Total raised now $132B. **Source:** Anthropic announcement, CNBC, TechCrunch **Significance:** Anthropic surpassed [[OpenAI]] as most valuable AI startup. $965B valuation implies near-trillion status at IPO. ### 2026-04-08 — D.C. Circuit denies stay of FASCSA designation **Category:** Government/Legal Three-member D.C. Circuit panel denied Anthropic's emergency motion to stay the "supply chain risk" designation. Court granted expedited briefing and set oral argument for May 19, 2026. Panel reasoned: "contained financial harm to a single company" vs. "judicial management of how the Department of War secures vital AI technology during active military conflict." **Source:** CNBC, legal analysis (KTS Law, Jones Walker, Mayer Brown) **Significance:** Split legal posture: Judge Lin's injunction blocks ban enforcement broadly, but D.C. Circuit allows FASCSA designation to stand for DoD contracts specifically. Anthropic excluded from defense work while other government use continues. ### 2026-02-27 — Government-wide ban; supply chain risk designation **Category:** Government/Legal Trump ordered government-wide ban on Anthropic products. [[Pete Hegseth]] designated Anthropic a "supply chain risk" under FASCSA. Trump publicly called the company "Leftwing nut jobs." Same evening, [[OpenAI]] announced its own classified Pentagon contract with no usage restrictions. **Source:** ABC News, NPR, CNBC **Significance:** The defining event. See [[Anthropic Government Ban]]. ### 2026-02-12 — $20M to Public First Action; Series G closes **Category:** Lobbying, Funding/IPO Anthropic donated $20M to Public First Action, a group supporting AI guardrails candidates in 30–50 state and federal races (both parties). Same day, Series G closed at $30B / $380B valuation. **Source:** CNBC, Anthropic announcement **Significance:** Largest AI-company political donation targeting safety regulation. Contrast: Anthropic gave $50K to Trump's inauguration (1/20th of what [[OpenAI]], Perplexity, and C3.ai each gave). ## See also - [[Anthropic]] — the main company page - [[Anthropic Dossier]] — the document collection front door - [[Anthropic Calendar]] — forward-looking beats - [[OpenAI Watch Log]] — the competitor's OSINT feed