A significant privacy enforcement action out of California: the California Attorney General and four District Attorneys have proposed a $12.75 million settlement with GM over the alleged sale and retention of drivers’ data. In this article, Stephanie Alvarez Salgado explains why the case matters—and why data minimization is emerging as a concrete compliance obligation, not merely a policy principle.

Who is the “volitional actor” when an AI system generates allegedly infringing output?

That question is central to ongoing copyright litigation against Anthropic and Perplexity. The platforms argue that users are responsible because outputs are generated in response to user prompts. Plaintiffs argue that platforms are responsible because they control training data, model architecture, and output constraints. Both positions draw on established copyright law, yet each captures only part of the causal chain.

The problem is that the volitional conduct doctrine was built for a different technological era: photocopiers, DVRs, and passive internet intermediaries where copying was discrete and attributable to a single actor. Generative AI does not work that way. There is also a growing tension in case law: if prompting may be insufficient to establish copyright authorship, can it nevertheless establish volitional conduct for direct liability?

In this new article, Chieh Tung explores why the traditional user-versus-system framework maps poorly onto AI systems where control is distributed across training, design, deployment, and prompting.

This week, the Supreme Court clarified the standard for contributory copyright infringement liability for service providers. In unanimously holding that knowledge alone is insufficient, but rather that liability requires intent shown through either inducement or provision of services tailored to infringement, the Court's decision could have far-reaching effects on the future of the contributory infringement analysis.  In this article, Ciara McHale and Jonathan Downing delve into this decision and what it may mean going forward.

Just last week, Hachette Book Group pulled a novel from distribution after allegations circulated widely online that large portions of the novel were AI-generated. Although Hachette has an AI disclosure policy for authors, and the author disputes having used AI, she states that an editor she hired may have. This scenario is a timely illustration of what's at stake when licensing AI-assisted works, which is the subject of Chieh Tung's third and final installment in the AI-assisted works and copyright series.

Part III of this series details best practices around commercializing AI-assisted assets and understanding the nuances around each stage of the creative chain. What rights and obligations do you have under the governing AI platform terms? What respective issues should licensors and licensees address when negotiating an AI-assisted work? And how should creators and owners think about how employees, works for hire, and service providers contribute to an AI-assisted work, and what implications does that have for the resulting work to be licensed?

Registering AI-assisted works is less complicated than it might seem. But the decisions around timing and disclosure have real consequences for what you can enforce and how.

Part II of Chieh Tung's series on copyright protection for AI-assisted works covers what content creators and owners need to know about registering AI-assisted works in the U.S., including what the Copyright Offices requires in terms of disclosures of AI-assisted works, how that disclosure affects the scope of a copyright claim, and practical guidance around timing and strategy.

AI is now standard in creative workflows. But when AI helps make the work, ownership and protectability become harder to assess.

In a new 3-part series, Chieh Tung examines how creators, companies, and content teams can better protect copyrights in AI-assisted works. Each installment includes a practical checklist for the people involved in creating, managing, and commercializing content. Part I addresses the legal framework surrounding human authorship and offers practical guidance on documenting creative workflows to better position those works for protection.

Check back next week for Part II, which will cover best practices for registering AI-assisted works.

The Ninth Circuit’s copyright framework is facing renewed scrutiny. In Sedlik v. Von Drachenberg, two concurring judges questioned whether the intrinsic test remains consistent with copyright law. A petition for rehearing en banc has now been filed, and the court has ordered a response.  In this insight, Chieh Tung summarizes the concurring opinions and the en banc petition and provides an overview of the most recent developments.

The next phase of AI litigation may turn less on copying and more on access controls and copyright management information under DMCA §§ 1201 and 1202.  In this insight, Chieh Tung examines the emerging decisions scrutinizing technical architecture and data ingestion practices and what they may mean for AI companies, platforms, and content owners.

Drawing on our recent wins at trial and on appeal before the Ninth Circuit, Tyz Law Group's resident Rhetorician Ciara McHale and Storyteller Jonathan Downing have joined forces to walk through the ins and outs of the extrinsic test's filtration framework.  Understanding filtration at all steps of litigation can be a copyright plaintiff's key to victory in an infringement claim, and we provide our battle-tested knowhow on the subject in our latest article.

The U.S. Supreme Court has agreed to hear a case that could significantly reshape the scope of liability under the Video Privacy Protection Act (VPPA), addressing who qualifies as a “consumer” under the statute. This decision may have meaningful implications for ongoing and future VPPA litigation nationwide.

Read our latest insight from Stephanie Alvarez Salgado for a breakdown of what’s at stake and what to watch next.