Anthropic News
Podcast 🎧 • Video 📽
Copyright & Legal
$1.5B Author Copyright Settlement Approved — Judge Signs Off on Largest Copyright Settlement in US History
US District Judge Araceli Martínez-Olguín granted final approval Monday of Anthropic's $1.5 billion settlement of a class action lawsuit brought by authors who accused the company of using their books without authorization to train Claude. The settlement is the largest known resolution of a copyright case in US history, and the first major settlement of the wave of AI copyright lawsuits that has moved through the courts since 2023.
Judge Martínez-Olguín rejected objections from some authors who argued the payout was too small.
The underlying training-data case produced a split ruling: a separate judge previously found that Anthropic had illegally downloaded and stored millions of copyrighted books, but ruled that training an AI model on copyrighted text constitutes fair use — a decision widely seen as a significant precedent for the AI industry.
Sources: Reuters · TechCrunch · Euronews · Claims Journal
Science & Health
Anthropic Opens Rare Disease Research Grants — Up to $50K in Claude Credits, Applications Through August 2
Anthropic announced a focused call for applications under its AI for Science program targeting rare genetic diseases, offering accepted researchers up to $50,000 in Claude credits over six months. The program runs two tracks: one for clinical researchers and patient organizations working on disease mechanisms, and one for early-stage biotechs accelerating drug development. Applications are open through August 2 at 11:59 PM PT.
Named early partners include the Monarch Initiative, an international consortium working on rare disease diagnosis, and Every Cure, which uses Claude to identify drug repurposing opportunities across millions of candidates.
The grants extend Anthropic's June 30 launch of Claude Science, an AI workbench for researchers, and its internal preclinical drug discovery program for neglected diseases — conditions that pharmaceutical companies typically pass over for economic reasons.
Accepted grantees access Claude Opus and other approved models via the API.
Sources: Anthropic · Clinical Research News · STAT News · ZME Science
Policy
White House Frontier Model Framework Nears Announcement — 30-Day Pre-Release Government Review Window for Anthropic, OpenAI, Google
The White House is finalizing a voluntary framework with Anthropic, OpenAI, and Google that would give federal agencies up to 30 days to review the national security implications of a new frontier model before public release. An announcement is expected before August 1, when a 60-day deadline from President Trump's June 2 executive order on AI and cybersecurity expires.
The evaluation benchmarks used to classify models are classified and run by the NSA. Meta is not part of the deal. The framework is voluntary and explicitly prohibits mandatory licensing or preclearance — language inserted to reassure industry that Washington was not building an approval regime.
Anthropic's participation is notable given that the company had a $200 million Pentagon contract terminated earlier in 2026 after refusing contract language permitting autonomous weapons use without human intervention, leading to a brief period in which the administration designated Anthropic a supply chain risk. Both sides have since resumed working relations.
Sources: Build Fast With AI · Eastern Herald · Twin Strata · Financial Times
Project Glasswing
90-Day Report Still Pending — 15 Days to Black Hat, White House Framework Adds Regulatory Context
Project Glasswing's 90-day progress report has not been published as of July 21. The program operates with approximately 200 partner organizations across 15 countries scanning critical infrastructure codebases with Claude Mythos 5.
Black Hat USA opens in 15 days on August 5 in Las Vegas. The imminent White House voluntary framework — which gives federal agencies a 30-day pre-release review window for frontier models — adds new regulatory context for how Glasswing's findings may be handled and disclosed.
Anthropic has made no announcement on the report.
Sources: Anthropic · Build Fast With AI · Decryption Digest
Claude Mythos
Mythos 5 Access Unchanged — White House Review Framework Would Apply to Any Future Broader Release
Claude Mythos 5 remains restricted to vetted Glasswing partners as of July 21, unchanged since the July 1 restoration. The White House voluntary framework nearing announcement this week is directly relevant: any future decision to expand Mythos access beyond the current Glasswing cohort would presumably fall under the 30-day pre-release government review window.
The framework's classified benchmarking process, run by the NSA, makes Mythos 5 a clear candidate for designation as a covered frontier model.
General availability carries no confirmed timeline.
Sources: Build Fast With AI · Eastern Herald · Anthropic