Intellectual Property Report

AI as a Tool, not a Legal Actor: How Courts Have Allocated Legal Responsibility for AI Across Different IP Regimes
Wendy Wang, Adam Nguyen (Law Clerk)
Intellectual property rights have historically existed in the context of human creators. The inventor, author, or holder of the works in dispute was always a human—or at the very least, a legal entity composed of humans. But what happens when the creator is not human or a human entity? What if the works at issue were created entirely by generative AI, such that inventorship, authorship, or ownership cannot be assigned to a human? As the capabilities of AI continue to advance, courts now face a new challenge: determining legal responsibility for AI-generated works.

Despite this recent technological shift, courts have declined to modify the principles of IP law to recognize AI as a creator of IP rights. Recent court decisions reflect a consistent underlying theme: AI is treated as a tool used by humans, not a legal actor capable of creating, owning, or bearing responsibility for IP rights or disputes.
Continue Reading

When Consumers Enforce Open Source: The SFC v. Vizio Case
Andrew Wilson, Jaemin Sung
A California court recently issued two important rulings in a lawsuit between a nonprofit organization and TV manufacturer Vizio over open source software. These decisions may affect how companies manage open source software compliance obligations and whether everyday consumers have standing to sue to enforce those obligations.
Continue Reading

Our Take on AI: May 2026
California Publishes Executive Order on AI: On March 30, 2026, Governor Gavin Newsom signed Executive Order N-5-26, expanding California's AI governance framework established by Executive Order N-12-23 (September 2023). The new order directs multiple state agencies to develop measures ensuring that Generative AI is procured and deployed responsibly across state government operations, with most actions carrying a 120-day deadline of July 28, 2026. Key provisions include new certification requirements for state contractors (covering content moderation, bias governance, and civil liberties protections), a federal supply chain risk review mechanism allowing California to diverge from federal designations it deems improper, contractor responsibility reforms targeting entities judicially determined to have undermined privacy or civil liberties, and forthcoming guidance on data minimization and AI watermarking aligned with California Business & Professional Code §§ 22757.2 and 22757.3. Although the order does not itself impose binding obligations on private entities, companies selling or licensing AI to California state agencies should begin preparing for new certification requirements and assessing their compliance posture ahead of the state's rulemaking process. You can read more about this development here: "California Publishes Executive Order on AI."

White House Publishes National Policy Framework for Artificial Intelligence: On March 20, 2026, the White House released its National Policy Framework for Artificial Intelligence, setting forth legislative recommendations to Congress across seven policy areas, including intellectual property rights and preemption of state AI regulation. The framework's most consequential IP statement is the Administration's view that training AI models on copyrighted material falls within the fair use exception, though it expressly recommends that Congress refrain from statutory action and allow the courts to resolve the fair use question. The framework also recommends that Congress enable licensing frameworks or collective rights systems permitting rights holders to collectively negotiate with AI providers without antitrust liability—a permissive rather than compulsory regime. Additionally, the framework proposes a federal right of publicity protecting individuals from unauthorized AI-generated digital replicas of their voice or likeness, with carve-outs for parody, satire, news reporting, and other First Amendment–protected expression. Taken together with the Administration's prior AI orders, the framework signals a "legislation-light" approach favoring industry-led standards and judicial resolution over prescriptive regulation. You can read more about this development here: "White House Publishes National Policy Framework for Artificial Intelligence."

GSA's New AI Clause—Major Changes for AI Procurement: On March 6, 2026, GSA published "Basic Safeguarding of Artificial Intelligence Systems," a proposed clause that would substantially rewrite the terms under which AI is sold to, or used in performance of work for, the federal government. The clause applies broadly to any contractor that provides AI to the government or uses AI tools internally during contract performance—capturing analytics platforms, coding assistants, and cybersecurity tools alike—and reaches upstream to service providers with no direct government contract. Four provisions are particularly impactful: (1) an order-of-precedence provision making GSAR 552.239-7001 supreme over conflicting commercial terms of service, coupled with a prohibition on AI systems refusing outputs based on "discretionary policies"; (2) an "American AI" requirement restricting AI components manufactured, developed, or controlled by non-U.S. entities, with no clear definition of "developed and produced"; (3) IP provisions assigning the government ownership of all data inputs, outputs, and "Custom Developments" (including modifications, enhancements, and workflows) that go well beyond standard FAR/DFARS data rights; and (4) an "Unbiased AI" mandate requiring truthful outputs free of "partisan or ideological judgments," enforced through undisclosed government benchmarking. OMB M-26-04 declares compliance "material to contract eligibility and payment"—language that tracks Universal Health Services v. Escobar, 579 U.S. 176 (2016), and could establish predicate materiality for False Claims Act liability. The clause also creates tension with the EU AI Act's safeguard requirements and state AI laws such as Colorado's AI Act. MAS holders would have 60 days to accept or lose schedule access once Refresh 31 issues. You can read more about this development here: "GSA's New AI Clause: Major Changes for AI Procurement."

EU Committee on Legal Affairs Issues Report on Generative AI and Copyright: On February 25, 2026, the European Parliament's Committee on Legal Affairs adopted an own-initiative report addressing the intersection of GenAI and copyright law. The report acknowledges widespread violation of copyright by GenAI providers—including unauthorized scraping, non-compliance with opt-out reservations, and use of pirated sources—while also recognizing that current EU copyright law, including the CDSM Directive (Directive 2019/790, amending Directives 96/9/EC and 2001/29/EC), is inadequate to address GenAI-specific licensing challenges. The report sets forth several significant calls to action, including: voluntary collective licensing agreements per sector to enable fair remuneration; transparency obligations requiring AI providers to itemize copyright-protected content used for training; designation of the EUIPO as the trusted intermediary for managing machine-readable opt-out registrations and sector-based licensing; and a rebuttable presumption that copyrighted works were used for training where transparency obligations are not met, with legal costs shifted to non-compliant AI providers. The Explanatory Statement accompanying the report suggests a flat-rate copyright fee for use of European creators' data potentially amounting to 5–7% of global turnover. The Parliament also affirms that EU copyright law applies to GenAI models placed on the EU market regardless of the jurisdiction where training occurred, and that non-compliant models should be barred from the EU market. Organizations operating in the EU should monitor developments and assess exposure to potential new licensing, transparency, and enforcement obligations. You can read more about this development here: "EU Committee on Legal Affairs Issues Report on Generative AI and Copyright."

AI Counsel Code Podcast
In the episode, "AI, Privacy, and Cybersecurity: What Companies Must Know Now," Maggie Welsh sits down with Michelle Molner to unpack how artificial intelligence is reshaping data privacy, cybersecurity risk, and regulatory enforcement. The conversation explores the growing tension between AI’s demand for large data sets and long‑standing privacy principles like data minimization, consent, and purpose limitation. Listen to the full episode here.

April 2026 Intellectual Property Report Recap
In case you missed it, here is a link to our April 2026 Intellectual Property Report.