BitLaw

AI Patent Litigation

Can AI be an Inventor?

A more detailed discussion of this issue can be found on Bitlaw's AI Inventors and Patent Applications web page.

Whether an artificial intelligence system can be named as an inventor on a patent is a key legal issue now being tested in courts. Thus far, courts and patent offices have uniformly held that only natural persons qualify as inventors under current law, rejecting attempts to name AI systems as inventors. However, these rulings leave unresolved the more complex and emerging question of how to handle inventions resulting from human-AI collaboration. As litigation in this area develops, it will force courts and lawmakers to confront how traditional concepts of inventorship apply when AI plays a substantive role in the inventive process.

Thaler v. Vidal (Fed. Cir. 2022)

The Federal Circuit affirmed the United States Patent and Trademark Office's (USPTO) denial of Stephen Thaler's patent applications because they did not list any human as an inventor. The core issue was whether an artificial intelligence system could be listed as an inventor on a patent application. Thaler, who developed AI systems that generated inventions, filed two patent applications, listing an AI system called DABUS as the sole inventor. The USPTO denied the applications, asserting that a machine cannot qualify as an inventor. The Federal Circuit upheld the USPTO's decision, asserting that under the Patent Act, an "inventor" must be a natural person, hence, AI systems cannot be recognized as inventors.

AI Copyright Litigation

Can AI be an Author?

A more detailed discussion of this issue can be found on Bitlaw's Copyright Protection for AI-Generated Works web page.

Both the Copyright Office and the Federal Courts have held that works created solely by artificial intelligence systems are not eligible for copyright protection, as authorship requires a human creator. However, the law remains unsettled when it comes to works generated through collaboration between humans and AI—particularly where the AI plays a significant creative role. As generative tools become more integrated into creative workflows, the question of how to handle the joint creation of works will become a key litigation issue.

Thaler v. Perlmutter (DC Cir. 2025)

The U.S. Court of Appeals for the D.C. Circuit affirmed that only works created by human beings are eligible for copyright protection under the Copyright Act of 1976. Stephen Thaler had attempted to register a copyright in an artwork titled "A Recent Entrance to Paradise,"" which he claimed was entirely generated by an AI system he created, known as the “Creativity Machine.” He identified the AI as the sole author and himself as the copyright claimant. The U.S. Copyright Office rejected the application, citing its longstanding policy that only works with human authorship are copyrightable. The district court upheld the denial. On appeal, this court held that the statutory term “author” in the Copyright Act refers exclusively to human beings. The court emphasized that the structure and language of the Act—including provisions relating to an author’s life span, heirs, nationality, and intent—implicitly require human authorship. The opinion also noted that the Copyright Office has consistently interpreted the statute this way since at least the 1970s, and that Congress has acquiesced in this interpretation. While acknowledging that humans may use AI tools to aid in creative work, the court drew a bright line: when a work is created autonomously by a machine, with no human creative input, it falls outside the scope of U.S. copyright protection.

Fair Use/Copyright Infringement while Training AI Models

A more detailed discussion of this issue can be found on Bitlaw's Fair Use and the Training of AI Models on Copyrighted Works web page.

Fair use has emerged as a central issue in the growing wave of AI-related litigation. Plaintiffs in multiple lawsuits allege that generative AI models unlawfully copied their copyrighted works during training, while defendants assert that such training qualifies as fair use under copyright law. Courts must now determine whether the unlicensed ingestion of copyrighted content for model development is legally permissible, and whether it constitutes transformative use or causes market harm. These cases will likely establish foundational precedents for how copyright law applies to machine learning and data-driven technologies.

Thomson Reuters Enterprise Centre GmbH v. Ross Intelligence Inc. (D.Del. 2025)

The court ruled in favor of Thomson Reuters, determining that Ross Intelligence infringed Thomson Reuters' copyrights by using protected editorial headnotes to train its AI-powered legal research tool. Ross indirectly accessed Westlaw’s proprietary headnotes through Bulk Memos compiled by a third-party provider (LegalEase), which Ross then utilized extensively to develop its AI system. Unlike generative AI systems that train on millions of documents to understand the connections between words to gain the ability to generate new documents, Ross's AI system trained on a small set of data based on Westlaw's headnotes to obtain the single ability to locate existing court cases based on user-submitted queries. Central to the court’s decision was the rejection of Ross’s fair use defense, which argued that the intermediate nature of its copying supported a finding of transformative use. The court found otherwise, emphasizing that Ross’s AI served exactly the same purpose as Westlaw’s headnotes—assisting users in identifying relevant judicial opinions. Additionally, the court distinguished this case from precedent involving computer software, noting that those cases involved functional source code where intermediate copying was necessary to access underlying non-copyrightable ideas. In contrast, the court underscored that Westlaw’s headnotes were expressive editorial content, not functional software, and thus intermediate copying did not justify a transformative fair use claim. Further, the court determined that Ross’s commercial use of the headnotes negatively impacted Thomson Reuters's market opportunities, further supporting the court’s conclusion that fair use did not apply.

In April, 2025, the trial judge certified two questions in this case for interlocutory appeal to the Third Circuit. First, whether the West headnotes and West Key Number System are original and therefore protected, copyrighted works. And second, whether Ross's use of the headnotes was fair use. This will shorten the time before we get the first appellate decision on the issue of whether training AI models on copyrighted works can constitute fair use.

Artificial Intelligence (AI) Patent Attorney

Please see Dan Tysver's bio and contact information if you need any AI-related legal assistance. Dan is a Minnesota-based attorney providing AI advice on intellectual property and litigation issues to clients across the country.