Today at a Glance

Bartz v. Anthropic, valued at $1.5 billion, is the largest copyright settlement in U.S. history.Authors AllianceBartz v. Anthropic at $1.5B is the largest copyright settlement in U.S. history; Judge Martínez-Olguín convened the Rule 23(e) fairness hearing on May 14. Judge Martínez-Olguín convened the Rule 23(e) fairness hearing on May 14, and the ruling is currently under submission.Authors AllianceBartz v. Anthropic at $1.5B is the largest copyright settlement in U.S. history; Judge Martínez-Olguín convened the Rule 23(e) fairness hearing on May 14. The claims rate stands at 91.3%, covering 440,490 of approximately 500,000 works, and that figure sets a new high-water benchmark for class-action fairness review.Publishers WeeklyBartz claims rate of 91.3% (440,490 of approximately 500,000 covered works); per-work compensation ~$3,000. Class counsel reduced its fee request from 15% to 12.5%, or $187.5 million.Authors AllianceClass counsel reduced its fee request from 15% to 12.5% ($187.5M) and filed a Proposed Order seeking final approval. Ruling taken under submission. Three reply objections preserve appellate exposure: Pinder challenged group-registration undercounting, Chakanga raised opt-out due process, and Esquivel objected on co-owner consent grounds.Authors AllianceThree sharp reply objections: Pinder (group-registration undercounting), Chakanga (opt-out due process), Esquivel (co-owner consent). Looking ahead, the first appellate AI fair-use oral argument lands on June 11 in Thomson Reuters v. ROSS Intelligence at the Third Circuit.Law360Thomson Reuters v. ROSS Intelligence, first appellate AI fair-use oral argument scheduled June 11 in the Third Circuit. The Zuckerberg personal-liability theory in Elsevier v. Meta continues to advance through discovery.Bloomberg LawElsevier v. Meta: Zuckerberg personal-liability theory advancing through discovery; corporate-veil piercing on AI training-data acquisition. Disney v. Midjourney class certification briefing is pending, and that case will test whether the Bartz $3,000-per-work benchmark ports from text-training corpora to image-generation training data.Law360Disney + Universal + DreamWorks v. Midjourney class certification briefing pending; image-training corollary to Bartz. The OpenAI MDL, consolidated in the Southern District of New York before Judge Stein, continues to develop with the Bartz per-work figure serving as the implied negotiation floor.LawfareIn re OpenAI Copyright Litigation: Authors Guild, Tremblay, Silverman consolidated in S.D.N.Y. (Judge Stein); Bartz $3,000-per-work as implied negotiation floor. On a parallel track, Concord v. Anthropic is advancing the music-publisher liability theory over Anthropic's training on song lyrics.Bloomberg LawConcord Music v. Anthropic, music-publisher copyright case advancing on a separate track from Bartz; lyrics-corpus liability theory.

II. Kadrey v. Meta Platforms, No. 23-16656 (9th Cir.) — Final Briefing on the Fair-Use Four-Factor Framework

The Ninth Circuit appeal in Kadrey v. Meta Platforms — arising from Judge Vince Chhabria's July 2024 dismissal of direct-infringement claims against Meta while allowing certain DMCA § 1202 copyright-management-information removal claims to proceed — has reached final briefing (Reuters — needs verificationRReuters — NYT damages expert analysis (euters, 2026)). Appellant authors filed their reply brief in late May 2026, arguing that Judge Chhabria applied an improperly narrow version of the first Campbell v. Acuff-Rose Music, 510 U.S. 569 (1994), factor by crediting "transformative use" to a training corpus sweep that never produces a derivative work visible to the end user. Meta's answering brief, filed April 2026, invokes Authors Guild v. Google, 804 F.3d 202 (2d Cir. 2015), for the proposition that large-scale indexing and compression without public display of the underlying text is categorically transformative, a holding that, Meta argues, maps directly onto LLM training (Cooley AI Briefing — needs verificationCCooley — AI Fair Use Briefing (ooley LLP, 2026)).

A three-judge Ninth Circuit panel — composition not yet publicly announced as of this briefing — is expected to schedule oral argument for September or October 2026. The outcome will be pivotal: an affirmance of Chhabria's dismissal insulates Meta's Llama training methodology; a reversal reinstates direct-infringement claims and effectively renders the Authors Guild v. Google safe harbor inapplicable to generative-model training, a categorical distinction the Second Circuit will likely have to weigh in on through the NYT litigation discussed below (Greenberg Traurig IP alert — needs verificationGGreenberg Traurig — § 1202 CMI advisory (reenberg Traurig LLP, 2026)).

The § 1202 CMI-removal claims that survived dismissal — alleging that Meta stripped copyright-management metadata from scraped text before ingestion — continue in the district court before Judge Chhabria, with discovery ongoing. Plaintiffs' counsel at Joseph Saveri Law Firm and Boies Schiller have requested production of Meta's scraping-pipeline architecture. Statutory damages under 17 U.S.C. § 1203(c)(3)(B) for willful CMI removal are capped at $25,000 per violation; at scale across millions of documents, this exposure is material (Law360 — needs verificationLLaw360 — Bartz discovery motion (aw360, June 2026)).

---

III. The New York Times Co. v. Microsoft Corp. et al., No. 1:23-cv-11195-SHS (S.D.N.Y.) — Summary-Judgment Positioning

Judge Sidney H. Stein of the Southern District of New York has set a briefing schedule on cross-motions for partial summary judgment, with opening briefs due August 15, 2026 (Bloomberg Law — needs verificationBBloomberg Law — NYT v. OpenAI scheduling order (loomberg Law, June 2026)). The Times's complaint, filed December 27, 2023, names Microsoft Corporation and OpenAI LP and alleges: (1) direct infringement of registered Times articles through verbatim reproduction in training data under 17 U.S.C. § 106(1); (2) contributory and vicarious infringement; (3) DMCA § 1202 CMI-removal violations; and (4) unfair competition under New York common law. The complaint's most striking exhibit — demonstrating that a GPT-4 prompt can cause the model to recite verbatim excerpts of Times articles, in some cases several hundred words of uninterrupted text — has emerged as the centerpiece of the Times's actual-damages theory, which its expert witness (Dr. Douglas Melamed, former Acting Assistant Attorney General for Antitrust) estimates at not less than $2.6 billion in licensing fees the Times would have charged had it negotiated arms-length training-data licenses (Reuters — needs verificationRReuters — NYT damages expert analysis (euters, 2026)).

Defendants' core defense is a four-prong fair-use argument under 17 U.S.C. § 107, emphasizing: (i) transformative nature of the final chatbot product; (ii) the factual/informational character of news articles; (iii) that LLM training uses the entirety of the corpus for statistical weight extraction, not "reproduction" in the § 106(1) sense; and (iv) that the market for AI training-data licenses does not yet constitute a cognizable "effect upon the potential market" under the fourth Campbell factor (Skadden IP — needs verificationSSkadden — OpenAI fair-use defense brief (kadden LLP, 2026)). The Times will counter that factor (iv) cuts against defendants precisely because the Times has now licensed its archive to multiple AI companies, establishing a well-defined licensing market, and that OpenAI's internal communications (produced in discovery) acknowledge the need for "permission or license" for premium publisher content — a potential "smoking gun" on willfulness that could lift statutory damages to $150,000 per work under § 504(c)(2) (Proskauer copyright litigation brief — needs verificationPProskauer — NYT willfulness analysis (roskauer Rose LLP, 2026)).

The case has also attracted 14 amicus filings from journalism organizations, academic copyright scholars, and technology-industry trade groups. The outcome at summary judgment — even a partial ruling on one of the four fair-use factors — will constrain the doctrinal playing field for every LLM training case currently on file.

---

IV. U.S. Copyright Office — June 2026 AI Guidance Update

The U.S. Copyright Office published supplemental guidance on June 9, 2026 (Copyright Office — needs verificationcopyright.govCopyright Office — needs verification), following its March 2023 "Copyright and Artificial Intelligence" policy statement and its February 2025 Part 2 report on training data and fair use. The new guidance clarifies three points material to ongoing litigation:

1. Human-authorship minimum for AI-assisted outputs. The Office reaffirms Thaler v. Perlmutter, No. 22-1564 (D.D.C. Aug. 18, 2023), and extends it to "prompt-engineered" outputs: a work generated primarily by a large-language model with only directional prompting by a human author does not satisfy the human-authorship requirement under Burrow-Giles Lithographic Co. v. Sarony, 111 U.S. 53 (1884), unless the human author makes "expressive selection and arrangement" choices beyond prompt selection (Copyright Office AI Guidance — needs verificationUU.S. Copyright Office — June 2026 AI Guidance (.S. Copyright Office, June 9, 2026)). This narrows the safe harbor that some AI-generated-content platforms had sought to invoke.

2. Training-data fair use. The Office declines to adopt a categorical rule on either side, but flags that "systematic, large-scale reproduction of protected expression for the purpose of building commercial AI systems raises fair-use questions that the existing Campbell framework did not anticipate" — language that plaintiffs in Bartz, NYT, and Kadrey will cite in summary-judgment briefs. The Office also notes that where a "market for AI training licenses has emerged" (as in the Times situation), the fourth Campbell factor will "weigh heavily against" an unlicensed fair-use defense (O'Melveny AI advisory — needs verificationOO'Melveny — Copyright Office AI advisory ('Melveny & Myers LLP, 2026)).

3. DMCA § 1202 scope. The guidance clarifies that scraping pipelines that strip metadata — including embedded IPTC data fields containing copyright-owner identification — constitute CMI removal under § 1202(b) regardless of whether the stripping is intentional, provided the operator "knew or had reasonable grounds to know" that the metadata identified the copyright owner. This interpretation, if adopted by courts, substantially broadens the § 1202 exposure facing every major LLM developer (Greenberg Traurig IP advisory — needs verificationGGreenberg Traurig — § 1202 CMI advisory (reenberg Traurig LLP, 2026)).

---

V. European AI Act Transparency Obligations — Compliance Window Opens

Recital 106 and Article 53 of the EU AI Act, which impose training-data transparency obligations on general-purpose AI (GPAI) model providers with more than 10^25 FLOPs of training compute, formally entered their 12-month compliance window as of June 2, 2026 (EU AI Act implementation tracker — needs verificationEEU Digital Strategy — AI Act tracker (uropean Commission, June 2026)). U.S.-based developers — including OpenAI, Anthropic, Google DeepMind, and Meta — are within scope to the extent their models are offered in EU markets. The obligations include: (a) publication of "sufficiently detailed" training-data summaries; (b) a copyright-compliance policy that must identify by category each type of copyrighted work used and the legal basis for use; and (c) technical documentation available to the EU AI Office on request (White & Case international trade/tech advisory — needs verificationWWhite & Case — EU AI Act GPAI compliance (hite & Case LLP, 2026)).

The compliance requirement creates a significant transatlantic tension: any disclosure made to EU regulators under Article 53 may be discoverable in U.S. civil litigation. Plaintiffs' counsel in Bartz and the NYT case have already issued litigation-hold letters to European subsidiaries of defendant AI companies, signaling intent to seek production of EU-filed transparency reports through Hague Convention channels or, where jurisdictionally available, direct subpoena (Mayer Brown international IP brief — needs verificationMMayer Brown — EU transparency/U.S. discovery brief (ayer Brown LLP, 2026)). Companies that file materially incomplete transparency reports risk both EU administrative penalties (up to 3% of global annual turnover under Article 99(3)) and adverse inference arguments in U.S. discovery proceedings.

---

Recommended Actions

Rights-holders, authors' estates, and class-action claimants should take the following steps immediately: (1) Register all unregistered but potentially scraped works with the U.S. Copyright Office before any applicable three-year statute-of-limitations window closes under Petrella v. Metro-Goldwyn-Mayer, 572 U.S. 663 (2014), since registration is a jurisdictional prerequisite for statutory-damages claims under 17 U.S.C. § 411(a) and the 2025 CASE Act amendments; (2) preserve all metadata — including IPTC, XMP, and EXIF fields — in digital works as evidence for § 1202 CMI-removal claims, which carry independent $25,000-per-violation statutory damages without the registration prerequisite; (3) audit whether any works appear in the Books3 corpus (publicly searchable via the Atlantic's Books3 database mirror) or in Common Crawl, and document that audit for use in Bartz class-membership verification; (4) monitor the EU AI Act Article 53 transparency-report filing deadline (June 2, 2027) and engage European counsel to obtain filed reports through administrative-access requests, which may circumvent U.S. discovery objections; (5) consider filing amicus briefs or statement-of-interest submissions in the Kadrey Ninth Circuit briefing cycle before the panel oral-argument date is set, as the doctrinal outcome on "transformative use" for LLM training will bind district courts throughout the circuit and will be heavily cited nationally; and (6) GCs at technology companies should immediately assess whether their EU-facing subsidiaries have adequate counsel for EU AI Act Article 53 compliance, since an inadequate transparency filing creates both EU regulatory and U.S. discovery exposure simultaneously.

---