Legal

Copyright law in the United States has always required human authorship. The Copyright Office has been explicit: works produced by machines without human creative input are not eligible for copyright protection. But the rise of AI code generation tools like GitHub Copilot creates a murkier middle ground.

When a developer writes a detailed prompt, reviews the output, selects certain suggestions over others, and integrates them into a larger system — how much human authorship is present? Enough to support copyright? The answer likely depends on the specificity of the prompt and the degree of creative judgment exercised.

The more immediate commercial question isn't about the output — it's about the training data. Copilot was trained on public GitHub repositories, many under licenses that require attribution or prohibit commercial use. Several class action lawsuits are working through courts arguing that this training constitutes infringement at scale.

The resolution will shape not just AI development tools, but the entire question of what "authorship" means when human creativity is mediated by increasingly capable machines. This is a doctrine in genuine flux, and the next five years of litigation will define it.

The Fourth Amendment protects against unreasonable searches and seizures. For most of American legal history, this meant physical spaces — homes, papers, effects. The digital age broke that framework in ways the doctrine has been struggling to repair ever since.

The third-party doctrine, established in the 1970s, held that information voluntarily shared with third parties carries no reasonable expectation of privacy. This made sense for bank records shared with a bank. It makes much less sense for the continuous stream of location data, communication metadata, and behavioral patterns that smartphones generate and transmit constantly — not through active disclosure, but simply by existing.

Carpenter v. United States (2018) was a significant step. The Supreme Court held that historical cell-site location information does require a warrant, carving out an exception to third-party doctrine for data that provides a comprehensive picture of a person's movements. Chief Justice Roberts' majority opinion acknowledged that the digital age requires rethinking old rules.

But Carpenter was narrow. It left open questions about real-time location data, other forms of digital metadata, and smart home devices. The courts are applying case-by-case reasoning to a surveillance infrastructure that operates at a scale the doctrine was never designed to handle.

A New York federal judge dismissed a sweeping copyright lawsuit in which a writer accused bestselling author Tracy Wolff of stealing her romance-fantasy book drafts to create the popular "Crave" series. The court issued a lengthy opinion after reviewing over 6,000 pages of manuscript drafts submitted as evidence.

The central holding: genre tropes common to romance-fantasy fiction — including "hot, sexy, dangerous boys," forbidden love, supernatural academies, and similar conventions — are not protected expression under copyright law. They are part of the genre's shared stock of ideas, available to any author.

Copyright protection attaches to specific creative expression, not to ideas, themes, or genre conventions. The court's analysis reinforced the long-established idea-expression dichotomy: you can copyright the specific words you write, but not the underlying tropes, plot structures, or character archetypes you employ.

For IP practitioners, the case is a useful reminder of the limits of copyright in genre fiction — and a warning about the evidentiary burden in substantial similarity claims, where volume of comparison material doesn't substitute for quality of protected expression.

The U.S. Patent and Trademark Office has released new guidance significantly expanding the scope of design patent protection to cover computer-generated images displayed via virtual reality, holograms, augmented reality, and similar technologies.

Previously, design patents for screen-based interfaces required the claimed design to appear on a specific article of manufacture — typically a screen or display. The new guidance moves beyond that constraint, allowing applicants to claim CGI designs that exist in projected or simulated environments without being tied to a particular physical display device.

IP attorneys have described this as a meaningful expansion from prior rules. As software interfaces, AR/VR environments, and holographic displays become commercially significant, the ability to obtain design patent protection for purely digital visual designs is increasingly valuable.

The guidance reflects the USPTO's recognition that the boundary between physical and digital products is dissolving — and that patent doctrine needs to keep pace with how users actually encounter and interact with protected designs in emerging technological contexts.

The Ninth Circuit Court of Appeals has issued a stay on a lower court order that banned Perplexity AI's shopping bot "Comet" from accessing Amazon's platform while an appeal is heard. The case raises unresolved questions about whether AI bots operating on behalf of human users can be blocked by platform terms of service, and what legal theories support such injunctions.

Comet is an AI agent that can autonomously browse and make purchases on Amazon on a user's behalf. Amazon took legal action arguing the bot violated its terms of service and potentially its IP rights by automating access to its systems at scale.

The Ninth Circuit's decision to pause the ban during appeal signals that the panel sees at least colorable arguments on Perplexity's side — though a stay is not a ruling on the merits. The outcome could have broad implications for the legality of AI agents that interact with third-party platforms.

This case sits at a complex intersection: platform IP and terms-of-service enforcement, AI agent rights, and the question of whether automated access constitutes a form of trespass to chattels or copyright infringement. It's one to watch closely as AI agent use expands.

A federal judge in the Northern District of California handed Apple a decisive win against Musi, a music streaming app with tens of millions of downloads that was removed from the App Store in September 2024. The ruling confirmed that Apple's Developer Program License Agreement grants it the right to remove any app "with or without cause" — as long as it provides the required notice.

Musi's service streamed music from YouTube without licensing deals with rights holders, arguing it was merely enhancing the user's own YouTube interactions. After YouTube raised intellectual property concerns, Apple removed the app. Musi sued, claiming Apple was contractually required to conduct a formal review before delisting.

Judge Eumi Lee rejected that interpretation. The DPLA's general termination right — "with or without cause" — was held to supersede more specific provisions that Musi argued required a reasonableness review. The court found those provisions didn't limit Apple's broad termination authority.

In a separate ruling, the judge sanctioned Musi's lawyers from Winston & Strawn for fabricating facts in their filings. The combination of dismissal with prejudice and attorney sanctions makes this a near-total loss for Musi. For developers, the ruling confirms that App Store agreements give Apple near-unchecked removal authority — a dynamic with significant implications for the ongoing debate over app store regulation.