top of page

Recommended Preliminary Guideline for Handling Artificial Intelligence Inventorship Concerns


The Leahy-Smith America Invents Act (AIA) does not address this question, and the USPTO requested comments from the public last year. Notably, the American Intellectual Property Law Association (AIPLA) responded and offered that "under current U.S. law, the inventor must be a natural person [and] that law should not be changed at this time." Specifically, the AIPLA's comment states:

There is not yet enough information available to know whether this is truly “inventive AI.” Moreover, if inventive AI does exist in the future, it will be necessary to consider what types of activities by AI entities would be considered as inventive contributions to the claimed invention. For example, it may be that “accidental” invention of something new by an AI entity trained to do something else (where the invention is an unintended byproduct of the intended function) should be treated differently than a product that an AI entity is trained to invent. Thus, until there are more examples of and more understanding around “inventive AI,” it is premature to consider a change to the existing law.

While, the AIPLA's suggestion to wait for more information before changing the law is reasonable, provisional rules must immediately be put into effect to begin the collection of this data to gain a better understanding. Where there is no human inventor, the USPTO has made it clear that it will not accept a patent application or process it. However, one can imagine that certain application will have multiple "contributors", and those contributors may include a human being and an artificial intelligence machine. In such cases, it may be prudent to let the human inventor apply for patent and allow a notation in the record to indicate the nature of the contribution by the artificial intelligence machine. This may be done simply by adding a checkbox to the Application Data Sheet asking whether there has been a contribution by AI and a statement of explanation as to the nature of that contribution. By doing this now, we may begin the process of collecting valuable data to fully understand the scope of the issue and lay the groundwork for new laws. Additionally, this simple change would not run afoul of current statutes or caselaw, but can only help the public better understand whether this issue is a practical problem or a purely academic exercise.

The AIPLA's suggestion is prudent, but analysis of this issue will continue to be premature until real-world data is collected.


bottom of page