The Duty to Disclose AI-Assisted Inventions

The New Guidance Creates Concerns

AI-assisted inventions require great care when complying with your Duty to Disclose to the USPTO. The new Inventor Guidance for AI-assisted Inventions holds an ominous warning.

However, special care should be taken by those individuals subject to this duty to ensure all material information is submitted to the USPTO to avoid any potential negative consequences.

The new Guidance explains that while AI-assisted inventions are not categorically unpatentable, the inventorship analysis should focus on human contribution. Furthermore, the Guidance requires that a natural person or persons must provide a significant contribution to all the claims in the application. If not, the claim isn’t patentable, and must be canceled. The focus of the analysis is on the natural human contribution. AI cannot be an inventor and cannot be named on the patent application.

The USPTO does not require an applicant to disclose whether AI was involved. This is interesting when compared to the US Copyright Office, which does require that any AI-assistance be disclosed. The USPTO is putting the burden on the applicant to determine if AI-assistance needs to be disclosed. The USPTO only wants to know when there is prima facie evidence that no natural human made a significant contribution to the invention claimed in one or more claim, i.e. whether a natural person’s involvement is material information to the examination of the application.

Every patent application requires that all inventors be correctly named. When AI assists in the invention, it becomes part of this analysis. Although AI cannot be named as an inventor, it still must be analyzed in the same manner as a natural human, as a co-inventor. Co-inventors of an invention are common. They may attribute to one claim, a few claims, or all the claims. When AI is involved in any claim, there must be a natural person who is a co-inventor of those claims.

If the analysis is clear, and a natural person is an inventor, no disclosure of AI needs to be made. If the analysis is not clear, then the relevant facts need to be disclosed to the USPTO, since this would be material to the patentability of the invention. Or the claims could be re-written to include a natural person’s significant contribution. The decision will revolve around the importance of the claim to the scope of the invention.

It’s common for patent applicants to disclose information, such as prior patents, that are not material, but only relevant. This is done to avoid the implication of inequitable conduct and avoid a violation of the duty of candor and good faith. The thought is to err on over disclosure. This may also be a tactic where AI is involved. An applicant might disclose the use of AI and the relevant facts to get the USPTO’s “seal of approval.”

The consequence of not disclosing material information can result in the patent being held unenforceable for inequitable conduct. Former Chief Judge Rader called it the “atomic bomb of patent law.” Although proving inequitable conduct is difficult, it can be an issue, and should be avoided. To prove inequitable conduct, it must be shown by clear and convincing evidence that the patentee: 1) knew of the information, 2) knew that it was material, and 3) made a deliberate decision to withhold it. With AI-assisted inventions, arguably, the only one of these three that would be an issue, is whether the information was material. Was there prima facie evidence that no natural human made a significant contribution to any claim in the patent? Making the decision that a natural human made a significant contribution, and that no information needs to be submitted, creates the risk that the patent will be held invalid. Therefore, unless it is clear, the wiser approach is to disclose, even though the hurdle to show inequitable conduct is very high. Better safe than sorry.


The Takeaway

The new Inventor Guidance for A.I.-assisted Inventions adds a new concern for patent applicants. The question of whether to disclose the involvement of AI and the relevant facts or to withhold it. It will likely be a topic of exploration in litigation involving AI-assisted patents and will add cost and uncertainty to these patents. This will be true regardless of whether you disclose or not. If you don’t the issue will be whether it was material and if you do, it will be whether you disclosed enough information. Also, since the new Guidance is retroactive, any existing patents will need to be reviewed to determine if there are claims that may be an issue from an inequitable conduct perspective and require further filings to remove the potential cloud on the patent.

Bill Honaker
Bill Honaker, The IP Guy

About the Author

Bill Honaker, “The IP Guy” is a former USPTO Examiner, a partner with Dickinson-Wright, and author of the forthcoming book, Invisible Assets – How to Maximize the Hidden Value in Your Business. To download a sample chapter, click here.

To get answers to your questions click here. To schedule a time to talk, you can access my calendar by clicking here, email, or call me at 248-433-7381.








Leave a comment