llms.blog
Anthropic

Anthropic trained Claude on 10 editions of a music industry bible. The author got $5,000.

Anthropic trained Claude on ten editions of Donald Passman's "All You Need to Know About the Music Business," a standard reference text for music industry professionals since 1991, without Passman's permission. A federal judge has now ruled that Passman cannot opt out of the $1.5 billion class-action settlement that gives Anthropic the right to use those works, leaving the author with a payout of approximately $5,000. The settlement, approved by U.S. District Judge Araceli Martinez-Olguin, reso

axc
axc
1 min

Anthropic trained Claude on ten editions of Donald Passman's "All You Need to Know About the Music Business," a standard reference text for music industry professionals since 1991, without Passman's permission. A federal judge has now ruled that Passman cannot opt out of the $1.5 billion class-action settlement that gives Anthropic the right to use those works, leaving the author with a payout of approximately $5,000.

The settlement, approved by U.S. District Judge Araceli Martinez-Olguin, resolves Bartz v. Anthropic PBC, a class action over Anthropic's use of millions of copyrighted books to train Claude without author consent. Eligible works receive between $1,200 and $1,500 each after a 50/50 split with publishers and deductions for legal fees, according to the Authors Guild. Anthropic's settlement portal listed four editions of Passman's book, totaling roughly $5,000.

Passman, represented by attorney Kenneth Freundlich, filed a late opt-out request arguing he never received direct notice of the settlement before the February 2026 deadline. Internal emails with his publisher, Simon and Schuster, revealed confusion about the proceedings. Passman said he was unaware of the settlement until after the deadline had passed.

Anthropic countered that Passman, a sophisticated attorney, failed to demonstrate "excusable neglect." The settlement administrator reported mailing multiple notices to Passman's known address, none returned as undeliverable, and noted that Passman had inquired about the settlement with his publisher as early as April 2026.

The judge denied the opt-out request, citing the administrator's records and the multiple notices sent. Passman remains bound to the settlement terms, granting Anthropic the right to use his books in exchange for the standard class payout.

The case highlights a tension at the heart of the settlement: smaller creators may benefit from class-action coverage, but higher-profile authors stand to lose significant legal and monetary leverage by being bound to blanket terms. Passman's book has sold more than 500,000 copies across ten-plus editions. Neither Passman nor his attorney has publicly commented on whether they plan further legal action.

Sources

Claude Trained on 10 Editions of Donald Passman's Music Industry Bible Without His Permission. His Payout from Anthropic? $5,000. - Digital Music News

Judge Refuses to Let Donald Passman Exit Anthropic Settlement - Digital Music News

What Authors Need to Know About the Anthropic Settlement - Authors Guild

AnthropicCopyrightAI Training
ShareX (Twitter)LinkedIn

Written by

axc

More to read

  • Forward KL vs. Reverse KL Divergence: Mode Covering, Mode Seeking, and the Alignment Dynamics of Large Language Models

    Forward KL vs. Reverse KL Divergence: Mode Covering, Mode Seeking, and the Alignment Dynamics of Large Language Models Every phase of modern large language model development, from multi-trillion token pre-training to reinforcement learning from human feedback (RLHF) and student model distillation, fundamentally revolves around minimizing statistical distance between probability distributions. The primary mathematical tool utilized for this purpose is the Kullback-Leibler (KL) divergence, introd

    1 min

  • Embedding Model Fine-Tuning in Production: Hard Negative Mining, Synthetic Data Pipelines, and Contrastive Distillation

    Production retrieval-augmented generation (RAG) and semantic search architectures frequently suffer from domain mismatch when relying on general-purpose embedding models. Off-the-shelf bi-encoders trained on broad web corpora often experience a 15% to 30% degradation in retrieval metrics such as NDCG@10 and MRR@10 when deployed on specialized enterprise corpora, including proprietary codebases, internal API schemas, clinical trials, and technical documentation. While downstream cross-encoder re

    1 min

  • The Gumbel-Softmax Trick: How Continuous Relaxations Enable Differentiable Discrete Sampling

    The Gumbel-Softmax Trick: How Continuous Relaxations Enable Differentiable Discrete Sampling In modern deep learning, end-to-end training depends on reverse-mode automatic differentiation. When an architecture operates on continuous tensors, computing gradients via the chain rule is straightforward. However, many foundational artificial intelligence problems involve discrete choices: selecting tokens from a fixed vocabulary, routing tokens to expert networks in a Mixture-of-Experts (MoE) archit

    1 min