With the advent of Chat GPT, artificial intelligence (AI) has entered the mainstream and become readily accessible. However, the intellectual property landscape around AI is far from settled and will likely evolve rapidly. This article provides suggestions for actions companies can take right now to ensure that their intellectual property procedures and practices are ready for these changes.

I. Patents – Inventorship and Implementation

For patents, AI can be involved in two ways: (1) AI is used to make the invention; or (2) AI is used to implement the invention. Depending on how AI is used, innovators need to take different steps to future prove their procedures.

A. AI used to make the invention.

In Thaler, the Federal Circuit held that AI cannot be the named inventor on the face of a patent because “the Patent Act requires an ‘inventor’ to be a natural person.” Thaler v. Vidal, 43 F. 4th 1207, 1208 (Fed. Cir. 2022), cert denied. While the Supreme Court denied certiorari, the Thaler decision was only a panel decision. Furthermore, the Federal Circuit reached its decision in Thaler based on interpreting the term “individual” in the Patent Act to be a human being. Id. at 1207. This leaves open the possibility for further changes in the case law pertaining to inventorship for patents involving use of AI in the inventive process.

1. Consider keeping a record of how AI is used.

For inventions in which AI is used to generate the invention, innovators may want to consider creating a detailed record of how AI was used to generate the invention. Such a detailed record could be part of the invention disclosure form. Having such a record will be helpful as the case law around Thaler develops.

2. Consider restricting use of public AI for innovation.

To the extent that AI is used to make the invention, access to the AI cannot be unrestricted. The use of public AI in the invention process has the potential of being a novelty-defeating disclosure. Furthermore, in view of Thaler, such disclosures would not fall under the exceptions to novelty-defeating disclosures in 35 U.S.C. § 102 made by the inventor. As such, companies should consider avoiding use of public AI as part of the innovation process.

3. Consider trade secret protection for AI-derived invention.

There may be circumstances where AI can legitimately be an inventor or sole inventor under the present definition of inventorship. Notwithstanding Thaler being a panel decision, it is unlikely that absent a statutory amendment to the Patent Act, AI will be the sole inventor. In situations where AI is the sole inventor, in view of Thaler, innovators may want to consider relying on trade secret protection instead to the extent that such protection is available.

B. AI used to implement the claimed invention.

AI can also be used to implement the claimed invention, i.e., the claimed invention includes the use of AI. Many patents to such inventions have already been granted. A simple search for artificial intelligence in claims identified over 7,000 different patent families, with the earliest claiming priority back to 2005. See, e.g., U.S. Patent No. 11,830,503, claiming priority to 2005 for a system for operating a virtual assistant in an automobile. By way of another example, U.S. Patent No. 11,7269,31 claims a method for artificial intelligence-enabled management of storage media access.

1. Consider monitoring granted AI applications to ensure applications have necessary standard language helpful for AI applications.

As with computer software patents, it is anticipated that the patent landscape around AI inventions will rapidly evolve. Innovators may want to consider monitoring granted patents in that technology space to devise successful claim strategies. Similar to computer readable media for software patents, innovators may also want to consider if standard disclosure paragraphs pertaining to AI need to be incorporated into patent filings.

II. Copyrights - Consider placing restrictions on how marketing uses AI to generate marketing materials.

The current policy of the U.S. copyright office is that AI cannot be a copyright owner, and human authorship is a prerequisite for a valid copyright to issue. See Copyright Registration Guidance on Works Containing Material Generated by Artificial Intelligence, 16190 Fed. Reg. 88 (Mar. 16, 2023). The U.S. District Court for the District of Columbia recently affirmed this policy. See Thaler v. Perlmutter, 2023 U.S.P.Q.2d 980 (D.DC. Aug. 18, 2023) (granting summary judgment in favor of Register of Copyrights on refusal to register copyright in work generated entirely by plaintiff Thaler’s “Creativity Machine” AI computer system). Nevertheless, AI remains as an incredibly useful tool to rapidly generate marketing materials such as brochures, tweets, or other content intended to be disseminated via the internet.

To the extent a company wants to have copyright protection for marketing materials, AI cannot be the sole source for the materials. Accordingly, companies may want to consider developing guidelines for their marketing departments as to how AI is used. For example, to retain copyright protection, it may be permissible to use AI to revise or tweak marketing materials initially prepared by humans. It may also be advisable to consider tracking which marketing materials were generated with AI, insofar as copyright protection may at least be claimed for the portions of works that were human-authored.

Using AI to generate marketing materials also presents an additional concern. If publicly available AI is used, it is possible that the AI may use third party copyrighted materials. Depending on the amount of such third party copyrighted materials, it is possible for the materials to include an amount of third party copyrighted materials beyond what some would consider fair use. As such, use of AI generated marketing materials also has the potential to being inadvertently accused of copyright infringement. For example, AI has been used to generate derivatives of pictures. There are several lawsuits pending that assert that such use violates copyrights. See, e.g. Andersen et al. v. Stability AI Ltd. et al., N.D. Cal. 3:23-cv-00201 (filed Jan. 13, 2023).

III. Trademarks – Considerations for using AI to ideate trademarks and create stylized designs.

Trademarks are used to associate a good or service with the source of such goods or services in the consumer’s eyes. As such, the use of AI to generate logos and the like, proves to be of least concern, and can even provide an end-run around the Copyright Office’s rules requiring human authorship for copyright protection, in that AI-authored logos can be registrable as trademarks so long as the applicant can validly claim ownership of the design. Nevertheless, companies should keep track of how AI is used ideate a trademark, and should further note that using AI to ideate potential marks does not necessarily indicate with any confidence whether such proposed marks are available for use and registration vis-à-vis third party registered and common law rights. Some trademark tools are becoming available to utilize AI for trademark clearance purposes, but such tools are still in an early stage and cannot be relied upon to avoid infringement.

IV. Conclusion

AI impacts patents, trademarks, and copyrights differently. Companies need to be mindful of AI’s impact and the constantly evolving law in each of these areas. For each of these types of IP, companies should consider reviewing their internal procedures to take increased access to AI into account.

This publication is intended for general information purposes only and does not and is not intended to constitute legal advice. The reader should consult with legal counsel to determine how laws or decisions discussed herein apply to the reader's specific circumstances.