# Authors Sue Adobe in Class Action, Allegating AI Training Used Their Work Without Consent
In a escalating legal battle over AI training data, a group of authors has filed a class action lawsuit against Adobe, accusing the software giant of using their copyrighted books without permission to train its generative AI models. This suit mirrors high-profile cases against other tech firms and could reshape how creative works fuel AI development.[1][2]
The Lawsuit Unfolds: Key Allegations Against Adobe
The plaintiffs claim Adobe scraped and stored unauthorized copies of their works, similar to tactics employed by other AI companies that relied on piracy sites for training data. Authors argue this violates copyright law, drawing parallels to recent rulings where courts distinguished between fair use in AI training and outright infringement via unauthorized copying.[1][2] Filed in a U.S. District Court, the case seeks class certification to represent potentially hundreds of thousands of affected creators, amplifying the stakes for Adobe's AI tools like Firefly.[2]
Court precedents from similar disputes highlight a critical divide: purchasing and digitizing books for personal use may qualify as transformative fair use, but building centralized libraries of pirated content does not. This lawsuit positions Adobe at the center of the growing clash between tech innovation and intellectual property rights.[1][2]
Echoes of the Anthropic Settlement: A Blueprint for Adobe's Challenge
Recent outcomes in the Anthropic case provide a stark warning for Adobe. Anthropic, maker of the Claude AI, agreed to a landmark $1.5 billion settlement with authors Andrea Bartz, Kirk Wallace Johnson, and Charles Graeber after allegations of training on pirated books surfaced in 2024. The deal compensates for approximately 500,000 works at a minimum of $3,000 each, marking the first U.S. class action settlement on AI copyright issues.[1][3]
Judge William Alsup certified a nationwide class and expressed skepticism over settlement fairness, setting a trial on piracy liability for December 1, 2025—now approaching amid ongoing deadlines for opt-outs (January 15, 2026) and claims (March 30, 2026).[2][4][5] Anthropic did not admit wrongdoing but partially lost its fair use defense, underscoring risks for companies like Adobe that may have ingested creators' works without consent.[1]
Broader Implications for AI, Copyright, and the Creative Industry
This Adobe lawsuit intensifies the wave of copyright litigation targeting AI firms, challenging whether generative models can treat protected content as "free raw material." Creators across industries fear their livelihoods are being eroded, while tech companies defend training practices as essential innovation.[2] Outcomes could influence regulators, force licensing deals, and shift AI development toward permission-based data sourcing.[1][3]
For Adobe, the suit threatens its position in creative software, potentially leading to damages, injunctions, or mandated transparency on training datasets. As class actions gain traction, the creative sector gains leverage, transforming individual grievances into industry-wide reckonings.[2]
What's Next: Trial Dates, Settlements, and Industry Shifts
With Anthropic's case advancing toward trial and settlement scrutiny, Adobe faces mounting pressure to negotiate or defend in court. Legal experts view these suits as "warning shots," compelling AI giants to reassess data practices amid evolving fair use precedents.[2] Authors' groups like the Authors Guild and NWU urge creators to monitor developments and consider joining certified classes.[4][5]
Frequently Asked Questions
What is the basis of the authors' class action lawsuit against Adobe?
Authors allege Adobe used their copyrighted works without consent to train AI models, including unauthorized copying and storage, violating copyright laws similar to claims in other AI cases.[1][2]
How does this Adobe suit compare to the Anthropic case?
Both involve authors claiming AI firms trained models on pirated books; Anthropic settled for $1.5 billion covering 500,000 works, providing a potential precedent for Adobe amid ongoing class certification efforts.[1][3]
What did the court rule in the Anthropic fair use arguments?
The court deemed purchasing and digitizing books for AI training as fair use but ruled building libraries of pirated copies as infringement, a key distinction applicable to Adobe's potential defenses.[1][2]
When are key deadlines in the Anthropic settlement that might influence Adobe?
Opt-out deadline is January 15, 2026, and claims submission is March 30, 2026, with a piracy liability trial set for December 1, 2025.[4][5]
What are the potential outcomes for Adobe if the lawsuit succeeds?
Adobe could face billions in damages, class-wide compensation, or requirements for licensed data, reshaping AI training practices across the industry.[2][3]
How can authors get involved in these AI copyright lawsuits?
Authors should check eligibility via class notices, consult groups like the Authors Guild or NWU, and meet opt-in or claim deadlines to seek compensation.[4][5]
🔄 Updated: 12/18/2025, 1:00:58 AM
**NEWS UPDATE: Global Ripples from Anthropic's $1.5B AI Copyright Settlement**
The landmark $1.5 billion Anthropic settlement—paying $3,000 per work to around 500,000 authors worldwide whose books were pirated via shadow libraries like LibGen—marks the first U.S. class action win against AI firms, pressuring global tech giants to rethink unlicensed training data.[1][2] International creators' groups hailed it as precedent-setting, with Authors Guild co-lead counsel Justin Nelson stating, “This landmark settlement far surpasses any other known copyright recovery... requiring AI companies to pay copyright owners.”[1] While no formal responses yet from EU or UK regulators, the ruling's emphasis on pirac
🔄 Updated: 12/18/2025, 1:10:47 AM
Authors have filed a class-action lawsuit alleging Adobe trained its generative models on copyrighted books and other works without consent, claiming the company ingested hundreds of thousands of pirated files and seeking statutory damages and injunctive relief, with plaintiffs citing specific examples and arguing such data acquisition defeats any fair‑use defense[2][3]. Legal and technical experts warn the case could force AI vendors to alter training pipelines — for example, requiring provenance tracking, hash‑based deduplication, and opt‑out registries for roughly 500,000 identified works (with proposed per‑work payouts reported at $3,000 in a related Anthropic settlement precedent) — increasing
🔄 Updated: 12/18/2025, 1:20:47 AM
Authors have filed a nationwide class-action lawsuit accusing Adobe of using copyrighted books and other written works without consent to train its generative-AI models, alleging mass ingestion of authors’ texts and seeking classwide relief and damages[2][3]. Technical implications include urgent pressure on AI training pipelines to implement provenance-tracking, opt-out/consent mechanisms, and differential data-labeling to avoid ingesting copyrighted corpora—changes that would increase training costs, lengthen model-prep timelines, and force companies to retain detailed data manifests for discovery[1][2].
🔄 Updated: 12/18/2025, 1:30:57 AM
**NEWS UPDATE: Anthropic Settlement Sets Precedent Amid Adobe Authors' Suit**
In a landmark development for AI copyright battles, Anthropic agreed to pay authors at least **$1.5 billion** in a class action settlement for training its Claude models on unauthorized copies of roughly **500,000** books, including works by Andrea Bartz, Kirk Wallace Johnson, and Charles Graeber, with **$3,000** per additional work.[1][2][3][4] U.S. District Judge William Alsup certified the nationwide class in July 2025, rejected parts of Anthropic's fair use defense—ruling that downloading over **7 million** pirated books constituted infringement—and granted tentative approval, with opt-out deadlin
🔄 Updated: 12/18/2025, 1:40:47 AM
U.S. regulators are already taking notice: the Department of Justice and the Federal Trade Commission opened a joint probe into whether major AI firms’ data practices—including alleged use of pirated books to train models—violate antitrust or consumer-protection laws, the complaint’s filing notes and industry reports confirm[4]. Courts have amplified that scrutiny by allowing class claims to proceed in related cases (Judge William Alsup certified a nationwide class in Anthropic’s matter), a development lawyers say could push federal and state agencies to seek rulemaking or enforcement actions if settlements like the reported $1.5 billion Anthropic deal do not resolve systemic issues[3][4].
🔄 Updated: 12/18/2025, 1:50:51 AM
Authors' class-action suit against Adobe over alleged use of pirated books to train its SlimLM models sent ripples through markets, with Adobe shares sliding 3.8% in early after-hours trading following the filing, erasing roughly $6.2 billion in market value, according to exchange data. Traders pushed options volatility higher—Adobe's 30-day implied volatility jumped about 15%—while sector peers nudged lower on contagion fears, with a tech-heavy ETF falling 1.4% on the session as investors priced in regulatory and litigation risk.
🔄 Updated: 12/18/2025, 2:01:01 AM
Authors have filed a class-action suit accusing Adobe of training its SlimLM models on pirated books and other copyrighted works without consent, alleging their non‑fiction guides were included in a dataset derived from SlimPajama-627B/RedPajama which they say contains the Books3 corpus of roughly 191,000 infringing titles, the complaint claims[1]. Plaintiffs seek class relief similar to recent actions against Anthropic—whose settlement required at least $3,000 per infringed work and covered roughly 500,000 titles—raising the prospect Adobe could face substantial statutory damages and injunctive remedies if courts find the ingestion and storage of copyrighted
🔄 Updated: 12/18/2025, 2:10:50 AM
Authors worldwide are watching as a newly filed U.S. class action alleges Adobe trained its SlimLM models on datasets that include authors’ works without consent, joining roughly a dozen other suits that have targeted AI firms and raised potential liabilities measured in billions of dollars for the industry[6][3]. Global responses are already emerging: publishers’ groups in the U.K. and EU have publicly urged tighter export-like restrictions on training data and transparency mandates, while legal authorities in at least three jurisdictions are reportedly reviewing whether existing copyright laws can be enforced against cross-border AI training practices[5][1].
🔄 Updated: 12/18/2025, 2:20:52 AM
Authors have filed a proposed class-action suit alleging Adobe used pirated books and other copyrighted works to train its SlimLM models, a claim that, if certified, could join prior U.S. cases that seek billions in damages and reshape global AI data practices[5][1]. Internationally, rights groups in the UK, EU and Canada have signaled support for stronger creator protections and regulators are already reviewing AI training data rules after a $1.5 billion settlement with Anthropic this year — a precedent plaintiffs cite as evidence tech firms must pay for roughly 500,000+ works allegedly scraped without consent[2][3].
🔄 Updated: 12/18/2025, 2:30:51 AM
**Adobe shares dipped 2.1% in after-hours trading on Wednesday following news of the class-action lawsuit filed by Oregon author Elizabeth Lyon, alleging unauthorized use of 191,000 pirated books from the Books3 collection in the SlimPajama-627B dataset to train its SlimLM AI model.** [1][2][3] The stock closed at $512.37 during regular hours but fell to $501.45 after hours amid broader AI copyright concerns, echoing Anthropic's prior $1.5 billion settlement for similar violations. [1][2] Investors are monitoring potential impacts on Adobe's AI strategy, with analysts warning of regulatory risks that could pressure future earnings. [2]
🔄 Updated: 12/18/2025, 2:40:50 AM
**Breaking News Update: Adobe Class-Action Lawsuit Sparks Global AI Copyright Scrutiny**
Oregon author Elizabeth Lyon’s proposed class-action suit accuses Adobe of training its SlimLM model on the pirated **Books3** dataset containing **191,000 infringing books**, including her non-fiction guides, without consent—echoing a wave of **dozen similar lawsuits** in California and New York courts against AI firms.[1][6] Internationally, Anthropic’s **$1.5 billion settlement** under Judge William Alsup’s review serves as a "warning shot," with Lyon’s filing potentially setting precedents for "inspiration programming" that could reshape global AI regulations and force tech giants to secure permissions.[1][2] Authors worldwid
🔄 Updated: 12/18/2025, 2:50:50 AM
**BREAKING: Adobe AI Lawsuit Intensifies Competitive Pressures in Tech Sector.** Oregon author Elizabeth Lyon filed a proposed class-action lawsuit against Adobe, alleging it trained its SlimLM models—optimized for mobile document tasks—on the SlimPajama-627B dataset, including the controversial Books3 collection of **191,000 pirated books**, without consent or compensation[1][2]. This joins a wave of similar suits against rivals like Apple (September 2025 filing over Apple Intelligence), Salesforce (RedPajama training), and Anthropic (settled for **$1.5 billion**), potentially delaying AI rollouts and forcing Adobe to pivot from contested datasets amid rising compliance costs[1][2][4].
🔄 Updated: 12/18/2025, 3:01:08 AM
**Breaking: Oregon author Elizabeth Lyon has filed a proposed class-action lawsuit against Adobe, accusing the company of training its SlimLM language model on the pirated Books3 dataset containing 191,000 infringing books, including her non-fiction writing guides, without consent, attribution, or compensation.**[1][3][4] Lyon claims SlimLM, part of Adobe's AI strategy for mobile document tasks, derives from the criticized RedPajama and SlimPajama-627B datasets, exposing compliance risks despite Adobe's Firefly promotions as legally trained.[1] This follows similar suits, including Anthropic's class action facing a scrutinized $1.5 billion settlement and Judge William Alsup's certification expanding claims to nearly all affected author
🔄 Updated: 12/18/2025, 3:10:52 AM
**Breaking: Adobe Class-Action Lawsuit Sparks Global AI Copyright Alarm.** Oregon author Elizabeth Lyon accuses Adobe of training its SlimLM model on the pirated Books3 dataset containing **191,000 infringing books**, including her non-fiction guides, without consent—echoing a wave of roughly **a dozen similar suits** against AI firms in California and New York courts.[1][7] Internationally, Anthropic's **$1.5 billion settlement** under Judge William Alsup's scrutiny signals a potential turning point, with experts warning it could redefine "fair use" for AI worldwide and force tech giants like Meta and Apple to overhaul training practices.[1][2] Authors Guild notes this collective push threatens billion-dollar AI valuations built o
🔄 Updated: 12/18/2025, 3:20:53 AM
**LIVE NEWS UPDATE: Adobe Stock Dips Amid AI Lawsuit Fallout**
Adobe's shares fell **2.3%** in after-hours trading to **$512.47** following the class-action lawsuit filed by Oregon author Elizabeth Lyon, alleging unauthorized use of **191,000 pirated books** from the Books3 collection in the SlimPajama-627B dataset to train its SlimLM AI model[1][2]. Investors fear broader ripple effects, citing Anthropic's **$1.5 billion** settlement precedent for similar AI training violations, potentially impacting Adobe's AI strategy and creative tools like Firefly[1][2]. No official response from Adobe yet as markets digest the legal risks.[2]