Who Qualifies as an Inventor in AI-Assisted Patent Claims?

AI-Assisted Patent Inventorship: Legal Requirements

Under current United States law, an inventor must be a natural person who contributes to the conception of an invention. An artificial intelligence system cannot be named as an inventor, even when it generates novel technical ideas or substantially completes the claimed design. However, a human may qualify as the inventor of an AI-assisted invention if that person contributes to its inventive concept. Prompting, selecting, and directing an AI are not automatically enough; the human contribution must amount to conception of at least part of the claimed invention. Different people may qualify as joint inventors for different aspects of a single invention, but each must make a significant inventive contribution.

Also worth reading: Documenting AI-Assisted Inventions: What Patent Review Requires? · What Does AI-Assisted Patent Inventorship Guidance Mean for Inventors? · How Should Patent Teams Govern AI-Assisted Drafting in 2026?

The critical distinction is between inventorship and patentability. Human involvement sufficient to establish inventorship does not, by itself, make an invention eligible for patent protection. Applicants must also satisfy statutory requirements, disclose the invention adequately, and explain any reliance on AI when reasonably necessary. USPTO guidance and related legal commentary emphasize that inventors should document their human contributions, review AI-generated outputs critically, and avoid claiming invention merely because they commissioned or operated the system. Section 256 procedures may permit correction of defective inventorship, but they cannot transform an AI into a legal inventor.

USPTO Guidance for AI-Generated Inventions

Under current USPTO guidance, an inventor must be a natural person who contributes significantly to the conception of a claimed invention. An artificial intelligence system cannot be named as an inventor, regardless of whether it generated a technical solution, designed an invention, or identified a novel patentable application. AI may assist with research, analysis, optimization, drafting, or experimentation, but those activities alone do not establish inventorship. For AI-assisted patent claims, the relevant inquiry focuses on the human contribution to the claimed features and their conception, not merely on the importance of the AI tool or the commercial value of its output.

Accordingly, patent applicants should identify each natural person who made a significant contribution to the conception of the claims and explain that person’s role using language tied to the specific claimed subject matter. Merely supplying a problem to an AI, requesting a result, or accepting its output may not be enough. Conversely, a human who contributes to the conception of a claimed invention may qualify even when AI performs much of the technical work, provided the human contribution is significant and is documented. Inventorship is only one requirement: the claimed invention must also satisfy statutory patentability requirements, including novelty, nonobviousness, utility, and adequate disclosure.

Human Contributions and Patent Eligibility

Under current United States patent law, an AI system cannot be named as an inventor. An inventor must be a natural person who contributes to the conception of the claimed invention. For AI-assisted inventions, the key question is not whether AI participated, but whether a human provided a significant inventive contribution. Generating code, identifying patterns, predicting outcomes, or automating routine steps generally does not, by itself, make a human an inventor if the person merely supplied an idea or directed the use of an existing tool. Conversely, selecting a technically meaningful solution, solving an inventive problem, or contributing to the conception of specific claims may qualify.

Inventorship is claim-specific. Different applicants may qualify for different claims depending on the human contributions to the claimed features, and a proper application should identify only the individuals who made qualifying contributions. Joint inventorship may apply when multiple humans contributed to the same claimed invention, even if their contributions were not identical. The USPTO’s guidance on AI-assisted inventions and the practical challenges discussed by patent practitioners, including correction mechanisms under Section 256, underscore the importance of documenting human contributions throughout development. Companies should preserve engineering records, laboratory notebooks, design decisions, and communications showing who conceived the claimed subject matter.

Correcting Inventorship Errors After Filing

Under U.S. patent law, an inventor must be a natural person who contributes to the conception of the claimed invention. AI systems, models, and software cannot be named inventors. The prosecution history must identify a human who made a significant intellectual contribution to the claimed features. Merely directing an AI, supplying a well-defined prompt, or selecting among outputs usually is not enough. Using AI as a tool while the human conceives the solution may leave the human as the proper inventor, but machine-generated features without a qualifying human contribution may require correction or render the claim impermissible.

Inventorship is therefore claim-specific. Naming too many contributors risks a §256 correction, while naming too few may leave AI-generated subject matter unsupported by a human inventor. Practitioners should preserve technical records, distinguish conception from routine implementation, and disclose AI use accurately. Before filing, applicants should review each claim against human contributions and consult current USPTO guidance; after an incorrect naming issue arises, corrective mechanisms may be limited, especially when uncorrectable inventorship cannot be remedied.

Global Strategies for AI Patent Claims

Under U.S. patent law, an inventor must be a natural person who contributes to the conception of the claimed invention. AI systems, unlike human inventors, cannot be named inventors or applicants. This is particularly important in AI-assisted inventions because courts have found that merely prompting, directing, or supervising an AI system is insufficient when the human does not contribute to the inventive concept. A qualifying inventor must contribute to the claimed solution, and the contribution must be reflected in the claims.

Patent owners should therefore preserve records showing how named inventors conceived the claimed features, including their use of AI tools, domain expertise, problem identification, selection of inputs, evaluation of outputs, and modification of proposed solutions. The “human element” analysis asks whether each claim was conceived by a natural person. This evidentiary discipline also supports correction efforts under Section 256 if inventorship errors arise. Strategically, companies should align claim drafting, inventor declarations, and internal documentation with human contribution rather than treating AI participation as inventorship.

AI-Inventor Rules Compared

Under current U.S. inventorship rules, only natural persons can be inventors, and their status depends on significant contributions to the claimed invention’s conception, rather than ownership, authorship, funding, or commercial value. AI tools may assist research, drafting, testing, or analysis, but they cannot themselves be named inventors. Patent inventorship also differs from copyright authorship: a human author ordinarily must satisfy copyright’s independent statutory requirements. Because a qualifying human contribution is essential, teams should preserve dated records showing who proposed each claimed feature, evaluated alternatives, and shaped the final conception. Failure to name a proper inventor can jeopardize enforceability, while erroneous naming may permit correction through applicable proceedings.