This year, the American Intellectual Property Law Association (AIPLA) hosted its Annual Meeting in downtown Washington, D.C. Though each session of the Meeting was devoted to a different substantive topic, nearly every session had one thing in common: a discussion about how artificial intelligence (AI) is affecting that area of intellectual property (IP) practice. AI-related discussions included "hot topics" in AI, AI-assisted inventorship, what copyright-friendly jurisdictions are best suited for training artificial intelligence models using copyrighted material, how to prosecute AI-related inventions when rejected under 35 U.S.C. § 101, among a long list of other topics relating to AI.
While the subject of AI in IP law is by no means new, the pervasive extent to which AI was discussed at the Meeting highlights that this topic remains far from settled and that new issues continue to arise.
During the session on "hot topics" in AI, it was noted that patent applications drafted with AI assistance cannot be filed in certain jurisdictions. Meanwhile, in another session focused on patent drafting basics, the speaker promoted the use of artificial intelligence in drafting patent applications. It was even mentioned during this session that an attendee's client is requiring their patent attorneys to use AI tools to aid in drafting their patent applications. This disparity highlights the challenges many IP firms are currently facing when it comes to using AI. Questions IP firms may consider prior to using AI tools to aid in drafting patent applications include the following. What jurisdictions does the client file in? What contracting states are the inventors either nationals or residents of? What AI-drafting tool is best suited for the firm? Is the same AI-drafting tool best suited for all of the firm's clients? How will the firm ensure that patent applications drafted with AI tools remain private and confidential during the iterative drafting process? What level of experience does an IP practitioner have to have before they can use AI-drafting tools? How can patent drafters review patent applications drafted using AI tools to ensure they are contextually and technically accurate and free of AI "hallucinations"?
Further, what parts of a patent application should practitioners allow AI-drafting tools to draft? Should AI-drafting tools be used to aid in the drafting of an entire patent application, or, perhaps, only the written description after a practitioner has drafted the claims and figures? While only allowing AI-drafting tools to aid in drafting the written description, within the context of the practitioner-drafted claims and figures, might seem reasonable, practitioners should be wary of this practice. The process of drafting the written description is often where the practitioner realizes the subtleties and nuances of the invention that may not have been previously appreciated while drafting the claims and figures. Those potentially important subtleties and nuances may be overlooked when using AI-drafting tools to draft the written description. This may have detrimental consequences that could affect the client-practitioner relationship and result in a weaker patent application that is challenging to prosecute. However, many practitioners consider AI-drafting tools as a means to reduce patent drafting costs. While that may be true, overall cost savings and application quality should be considered, because relying on AI tools may result in additional costs being incurred during prosecution and in post-grant proceedings or, worse, result in abandoned applications that may have been patentable. Taking a more global perspective, missing the subtleties and nuances of an invention within a patent application could weaken a client's patent portfolio over time depending on their IP goals.
Because the use of AI in IP is here or, at least, inevitable, capturing the subtleties and nuances of the invention while using AI-drafting tools may benefit from a reorganization in how inventions are disseminated from clients to practitioners prior to drafting. Short and brief written information disclosures detailing the invention may no longer be sufficient. To supplement written information disclosures, clients might consider required information disclosure meetings or, if too time intensive, information disclosure videos.
While the meeting concluded with many questions left unanswered, the raising of these questions and discussions thereof highlights the importance of IP-friendly communication and collaboration between clients, practitioners, firms, and jurisdictions. Because AI in IP will most likely never have universally-applicable solutions, it is imperative that clients, practitioners, firms, and jurisdictions maintain open lines of communication surrounding AI-related topics. Only by open and collegial communication can the relevant stakeholders find what works best for them considering their differing IP goals. To say the AIPLA Annual Meeting fostered this friendly communication would be an understatement. Accordingly, it is important as IP practitioners to attend IP-related events to continue the AI discussions that are far from over and ever changing.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
