What Does Revised USPTO AI Inventorship Guidance Mean for Inventors?

USPTO's Revised AI Inventorship Framework

The revised USPTO AI inventorship guidance clarifies that an AI system cannot itself be named as an inventor under current U.S. patent law, but a natural person may qualify when their contribution to an AI-assisted invention goes beyond merely prompting the machine or accepting its output. The key question is not whether AI was involved, but whether a human made a significant contribution to the claimed invention. This approach, discussed by Patent Review Pro alongside the USPTO’s materials and commentary from Holland & Knight and Morgan Lewis, preserves the statutory human-inventorship requirement while recognizing modern inventive workflows.

Also worth reading: How Should AI Patent Eligibility Strategy 2026 Adapt to USPTO Guidance? · How Does AI-Assisted Invention Patentability Work Under New USPTO Guidance? · How Should Practitioners Use the USPTO’s New AI Search Guidance in 2026?

For inventors, the guidance means documentation should explain how they conceived, planned, or otherwise shaped the claimed subject matter, including material selections made with AI assistance. Commentary from Jones Day and IPWatchdog highlights the revision’s rejection of the approach advanced in Thaler v. Vidal and its emphasis on the human element. Inventors and patent practitioners should therefore review claims, assess each contributor, and preserve records showing the nature and significance of human input.

The USPTO’s revised guidance clarifies that AI cannot be named as an inventor under U.S. patent law. Inventorship remains a legal question centered on the natural person or persons who contributed to the claimed invention. AI tools may assist with drafting, searching, analyzing, or even proposing technical solutions, but the application must still identify the humans responsible for the inventive concepts described in the claims. This is especially important when AI-generated suggestions are incorporated without meaningful human evaluation or modification.

For inventors, the guidance highlights the need to document how AI tools were used and who made the final inventive decisions. Claim-by-claim analysis remains essential because a person need not contribute to every feature, but each inventor must have contributed significantly to at least one claim. Inventors should preserve relevant prompts, outputs, notes, experiments, and design records, while avoiding disclosure of confidential information and unsupported assertions that a human alone conceived every element. The revision provides greater practical clarity about AI-assisted work, but it does not create a new category of AI inventorship or replace the USPTO’s demand for a correct human inventor oath.

AI Contributions and Inventor Identifications

The USPTO’s revised AI inventorship guidance clarifies that an invention must still have a human inventor, while also requiring patent applications to identify the human who contributed to each claimed feature. AI-generated output cannot itself qualify as an inventor, and merely using an AI system does not make its user the inventor of every resulting claim. The relevant inquiry focuses on the contribution of each named inventor to the subject matter claimed, including specific ideas, arrangements, or improvements. This guidance is significant because it provides a more practical framework for AI-assisted inventions without abandoning the requirement that inventorship be a legal and human attribution.

For inventors, the guidance means that careful documentation of AI involvement is essential before filing. Applicants should preserve prompts, outputs, experiments, design decisions, and records showing which human person supplied the inventive concept or modified the AI-generated material. The guidance also reflects the USPTO’s rejection of the approach in Thaler v. Vidal, which sought to name an AI as the sole inventor. Inventors using AI should therefore disclose the appropriate human contributors, avoid unsupported ownership of purely machine-generated ideas, and consult patent counsel when determining who should be named for each claim.

Practical Filing and Prosecution Considerations

The revised USPTO AI inventorship guidance means that an AI system cannot be named as an inventor, even when it generates a significant portion of a claimed invention. Patent applications must identify natural persons who made a significant contribution to the claimed subject matter. Inventorship is assessed based on the conception of the invention, not merely on a person’s possession of a valuable idea or commercial insight. Accordingly, inventors should document how AI tools were used and what they independently contributed to the claimed features. Under the guidance, merely prompting an AI or requesting a particular result generally may not constitute inventorship. However, humans who significantly contribute to the inventive concepts remain eligible to be named as inventors.

For practitioners, the guidance places renewed emphasis on careful claim analysis during prosecution. Disclosures should explain the role of both human contributors and AI tools, while applications should avoid unsupported assertions that a human conceived an invention that was actually supplied by an AI. Prosecutors may also consider how a claim is framed and whether its limitations are attributable to a qualifying human inventor. A disciplined prosecution strategy therefore requires preserving development records, drafting applications around genuine human contribution, and avoiding the assumption that a person who conceived an invention through AI use is necessarily an inventor.

Thaler and the Limits of AI Autorship

The USPTO’s revised guidance clarifies that AI-assisted inventions may qualify for patent protection when natural persons contribute to their conception. Using AI does not automatically disqualify an applicant from inventorship, nor does it automatically establish a human’s inventorship. The decisive question is whether a person made a significant contribution to the claimed invention, including to specific claims or elements. Merely owning, operating, or directing an AI system generally is insufficient. Because AI cannot be named an inventor, applicants must identify qualifying human contributors and explain their roles through the required inventor declaration.

The guidance also narrows the implications of Thaler v. Vidal. While Thaler confirms that a human inventor is required, the USPTO’s position is not that any use of AI bars patent eligibility. Instead, patent offices must examine the actual human contribution, while courts retain authority to determine whether AI-generated material constitutes patentable subject matter. Revised guidance was intended to provide greater consistency, but it cannot resolve the unresolved statutory question of whether inventions made substantially by AI satisfy the patentability requirements. Applicants should therefore document meaningful human conception, disclose material AI use, and avoid attributing inventorship to the system itself.

AI Inventorship Guidance: Before and After

The revised USPTO guidance preserves the human inventor as the legal source of patent inventorship while clarifying that AI assistance is not categorically disqualifying. For each claim, the focus remains whether a natural person made a significant contribution to the claimed invention, considering the claim as a whole. Inventors should therefore document their inventive work, carefully describe the role of AI tools, and coordinate with counsel when drafting applications to avoid uncertainty over conception, contribution, and inventorship.