USPTO Seeks Public Comment on the Impact of Artificial Intelligence on Patentability

A Crossroads of Innovation: Unveiling the Challenges and Opportunities of AI-Driven Invention

The relentless march of technological progress presents a unique challenge for the intellectual property (IP) landscape. Earlier this year, the United States Patent and Trademark Office (USPTO) issued a Request for Comments (RFC) on inventorship guidance for AI-assisted inventions. See 89 FR 10043, available here (last visited June 10, 2024). Many responses to that inquiry, which closes on June 20, 2024, have encouraged the USPTO to investigate how AI impacts obviousness determinations. Not surprisingly, the USPTO recently issued a RFC seeking public input on the potential impact of Artificial Intelligence (AI) on prior art, the knowledge of a person having ordinary skill in the art (PHOSITA), and determinations of patentability. See 89 FR 34217, p. 34217, available here (last visited June 10, 2024). This client alert summarizes that RFC and delves into the complexities surrounding AI and patents, exploring the implications for patent applications, patent owners, patent practitioners, and the future of IP law.

Continuing the Conversation: Contributing to Future Policy
The RFC poses 15 questions related to the impact of AI on assessing prior art, the knowledge of a PHOSITA, and patentability. Prefacing these questions is a discussion of the USPTO’s position on the types of references available as prior art under 35 USC 102 and 103 and the case law followed by the USPTO when assessing a claimed invention’s nonobviousness under 35 USC 103.

• Prior Art: The USPTO deems a reference to be a “printed publication” available as prior art when it is “publicly accessible.” Id. at 34218. Additionally, the USPTO presumes that the disclosure of a reference is “presumed to be operable” (i.e., “provides a description that enables the public to make and use the disclosure”). Id. However, the USPTO acknowledges that it does not currently consider who (or what) made the disclosure. Id. at 34219. Because the USPTO recognizes that “AI may be used to create vast numbers of disclosures that may have been generated without any human contribution, supervision, or review,” the USPTO seeks input on questions about “whether AI-generated disclosures (that have not been prepared and reviewed by a human) should be afforded the same rebuttable presumption that they are operable and enabled.” Id.
• Nonobviousness: When assessing the nonobviousness of a claimed invention as required by 35 USC 103, the USPTO follows the test articulated by the Supreme Court in Graham v. John Deere Co. and KSR International Co. v. Teleflex Inc. Because neither 35 USC 103 nor case law explicitly requires that the “person” in the PHOSITA assessment be a human, the USPTO seeks input on questions about how AI impacts the assessment a PHOSITA’s level of ordinary skill in the art, if at all.

The USPTO encourages all interested parties to submit comments by the deadline of July 29, 2024, which will be announced on their website (https://www.uspto.gov/). Responses should address the questions posed in the RFC, and should also include a discussion of key areas like:

• Specific Case Studies: Detailing concrete examples of how AI is used in certain fields and the challenges faced in securing patent protection for AI-driven inventions will add valuable context.
• Benefits and Risks: Highlighting the potential benefits of AI for innovation while acknowledging potential risks, such as concerns about ownership and ethical considerations surrounding AI-generated inventions. Propose solutions for how existing patent law can adapt to accommodate the complexities of AI-assisted inventions.
• Policy Recommendations: Offering specific policy recommendations for the USPTO and Congress to consider when formulating guidelines and legislation on AI and patentability.

Guidance and Legislation: A Collaborative Approach to Shaping the Future
The USPTO’s prior RFC on inventorship guidance for AI-assisted inventions and current RFC on the impact of AI on assessing prior art, a PHOSITA’s level of knowledge, and patentability are forward-thinking steps. They recognize the transformative potential of AI and the potential need for updated frameworks to ensure a robust patent system. This collaborative approach underscores the multifaceted nature of integrating AI into the patent system, encompassing:

• The Shifting Landscape of Prior Art: A cornerstone of patent law is the concept of “prior art” – publicly available information that can be used to assess novelty and nonobviousness. The USPTO is grappling with how to navigate AI-generated outputs in this context.
• The Leveling Effect of AI: The increasing sophistication of AI tools might alter the standard for the “level of ordinary skill in the art.” This hypothetical person represents the baseline knowledge and skill within a particular technological field. Will AI advancements lower the bar for what constitutes an inventive step?

A Multifaceted Impact: Stakeholders Brace for Change
The USPTO’s current inquiry suggests that the USPTO may issue guidance on how to treat AI-generated prior art and assess the knowledge of a PHOSITA, which has the potential to impact various stakeholders in the innovation ecosystem:

• Patent Applicants: AI’s ability to generate prior art references quickly could pose initial obstacles to an applicant’s ability to obtain a patent. With this in mind, a patent applicant should consider drafting its application to include working examples and empirical data based on experiments and claims that include additional limitations to improve its ability to overcome AI-generated disclosures should they be treated as prior art by the USPTO. The cost in terms of time and money to obtain patents could increase if AI-generated disclosures are available as references because of lengthened prosecution times and difficulty in arguing novelty or nonobviousness. Malicious actors could also potentially use AI to generate prior art specifically targeted to competing technology, which can cause thickets and block efforts to obtain intellectual property by creating invalidating prior art.
• Patent Owners: AI’s ability to generate prior art references could increase the amount of prior art available to challenge a patent owner’s patent.
  Patent owners may want to analyze AI-generated references carefully to assess their public availability date.
• Patent Practitioners: AI can generate massive amounts of text and designs, potentially creating a flood of prior art for examiners to consider. This could lengthen prosecution times and make it harder for prosecutors to argue for the novelty or nonobviousness of an invention.

Looking Ahead: Charting a Course for the Future of AI and Patents
The USPTO will review public comments and use them to develop a comprehensive approach to AI and patentability. While the exact timeline for further developments from the Office remains uncertain, stakeholders should expect the USPTO to issue additional guidance or take regulatory actions in the future.

Staying informed of these developments and actively participating in the ongoing dialogue will be crucial for stakeholders who are tasked with navigating this continually evolving landscape. As the interplay between AI and patent law continues to take shape, a collaborative effort will be instrumental in fostering a future where AI can unlock boundless innovation while upholding the principles of a robust and adaptable IP system.

Our AI team is following updates on law and policy impacting intellectual property in the AI space and will keep our subscribers informed of new developments.