Nine plaintiff organizations now have active suits against Perplexity AI as of May 31; OpenAI's Frontier Governance Framework sets the EU AI Act compliance baseline for all frontier labs; Bartz v. Anthropic final approval order still pending with June 11 distribution-calculation date at risk.
As of May 31, 2026, nine media and publishing organizations have active copyright suits against Perplexity AI in a coordinated multi-front campaign that now constitutes the broadest plaintiff coalition assembled against any single AI defendant in the current litigation wave. CNN's May 28 filing in the Southern District of New York — alleging unlawful scraping and redistribution of more than 17,000 news assets alongside trademark and unfair-competition claims — is the most recent domestic action, joining the already-pending complaints from News Corp and Dow Jones, the New York Post, the Chicago Tribune, Encyclopedia Britannica, Merriam-Webster, and Reddit, all clustered in SDNY. Japan's Yomiuri Shimbun, which filed its own action in Tokyo District Court in August 2025 alleging unauthorized access to 119,467 articles and seeking nearly $15 million in damages, anchors the international dimension of the litigation. The pattern is structurally identical to the Concord Music aggregation theory: rights-holder coalitions filing serially in the same district to consolidate fair-use rulings, amplify statutory-damages exposure across nine complaint dockets, and force a global licensing resolution before any single court reaches the merits. (TechTimes) (CNN Business) (Nieman Lab — Yomiuri)
OpenAI's publication of its Frontier Governance Framework on May 28, 2026 — a document explicitly calibrating the company's safety and preparedness practices against the EU AI Act's Code of Practice for General Purpose AI and California's Transparency in Frontier AI Act — marks the most consequential unilateral compliance posture by any frontier lab ahead of the EU Act's August 2, 2026 enforcement date. The framework's relevance to the copyright dockets is direct: Articles 53(1)(c) and (d) of the AI Act require GPAI model providers to establish a policy to identify and comply with rights reservations expressed under the EU DSM Directive and to publish "a sufficiently detailed summary" of training contents. By publishing ahead of the enforcement deadline, OpenAI establishes a public reference baseline against which Anthropic, Google DeepMind, and xAI — all defendants in active U.S. copyright proceedings — are now implicitly expected to match. Any material gap between OpenAI's disclosed compliance posture and a competitor's public disclosures is likely to become a discovery argument and, ultimately, a § 107(4) market-harm exhibit in pending fair-use briefings, where plaintiffs will argue that the failure to license was commercially unreasonable given available frameworks and a hardening licensing market now documented by deals including News Corp–Meta ($50M/year, three-year term) and 17 publishers transacting through Snowflake's AI licensing platform as of May 2026. (OpenAI Frontier Governance Framework) (TechTimes) (Digiday — Snowflake Platform)
The Bartz v. Anthropic docket continues its post-hearing quiet, and the silence is now becoming consequential: as of June 1, 2026, no final approval order, no attorneys'-fees ruling, and no decision on the disputed late opt-outs has issued from Judge Martínez-Olguín — eighteen days after the May 14 fairness hearing and eleven days after class counsel and Anthropic filed their supplemental briefs on the late-opt-out question by the court's May 21 deadline. The settlement administrator's June 11 distribution-calculation date, at which point the net amount available for distribution (roughly $1.29 billion after the reduced $187.5 million fee petition) would crystallize, is now at material risk absent a court order in the next ten days. Eight of the original 53 objectors have retained appellate counsel, keeping the Ninth Circuit's Bartz, et al. v. Anthropic, PBC, No. 25-4843 appellate docket as a live contingency even after final approval; the group-registration collapse, publisher-favoritism, and statutory-damages-adequacy challenges are all fully briefed and ready for that track. (AI Lawsuit Tracker) (Anthropic Settlement Key Dates) (Authors Alliance — Settlement FAQ)
The most directly consequential dispositive-motion deadline of the current cycle arrives this week: cross-motions on fair use of lyrics as AI training data in Concord Music Group v. Anthropic (5:24-cv-03811, N.D. Cal., Judge Eumi Lee) are fully due June 8, 2026, placing before the court the first post-Bartz application of the § 107 four-factor framework to a training-data case outside the literary context. The simultaneous pendency of Anthropic's motion to stay Concord II (5:26-cv-00880) — heard May 20, still under submission — leaves open whether Judge Lee will have to adjudicate the $3.1 billion follow-on complaint covering 20,000-plus additional musical works at the same time as the original fair-use cross-motions. The Concord I briefing, combined with the CJEU's impending Advocate General opinion in Like Company v. Google (expected September 3, 2026), will provide the first coordinated EU–U.S. doctrinal framework for AI training fair use, and each Concord ruling will directly supply or undermine the appellate record available to the eight Bartz objector groups pressing the statutory-damages-adequacy argument. (BakerHostetler — Concord) (CourtListener — Concord I) (Bird & Bird — CJEU)
Rights-holders and broadcaster counsel tracking the Perplexity multi-plaintiff front should note that the coalition has now reached nine plaintiffs spanning three jurisdictions — SDNY, UK, and Tokyo — a concentration that maximizes both statutory-damages exposure and the collateral estoppel value of any adverse SDNY fair-use ruling for the Tokyo and UK proceedings; the CNN v. Perplexity complaint remains the cleanest pleading template for broadcasters and news publishers evaluating whether to join the front, particularly its trademark and unfair-competition counts built around the Comet Plus paid-tier misrepresentation. For AI developers and their E&O carriers, OpenAI's May 28 Frontier Governance Framework now defines the industry baseline for EU AI Act training-data compliance and § 107(4) market-harm defense: any GPAI provider that cannot publish an equivalent framework before August 2, 2026 faces both regulatory exposure under the Act and a courtroom disadvantage in fair-use briefing where the hardening licensing market (News Corp–Meta, Snowflake publisher platform, SPUR coalition) will be offered as proof that commercially reasonable licensing alternatives exist. Bartz class members should treat the June 11 distribution-calculation date as aspirational and verify claim status at anthropiccopyrightsettlement.com; the court cannot begin the distribution process until the final approval order issues, and any delay by the court this week pushes payment timelines into July and beyond — late-opt-out applicants should anticipate denial absent extraordinary circumstances given the supplemental briefing record. Practitioners advising clients on Concord I should incorporate the growing licensing-deal comparator set into their June 8 briefings: any submission that omits the News Corp–Meta and Snowflake platform data as § 107(4) market-harm evidence is leaving the strongest available record on the table.
---
This advisory is provided for informational purposes and does not constitute legal advice.For assessment of class membership, claim eligibility, or litigation exposure, consult qualified counsel.
AI Copyright, Bartz, Anthropic, Perplexity, Advisory