Artificial intelligence tools like ChatGPT have revolutionized how we approach research and problem-solving. From answering complex queries to drafting content, they have become integral to modern workflows—including patent research. But here’s the million-dollar question: Does searching AI tools like ChatGPT constitute a public disclosure that might impact the patentability of an invention?
If you’re a patent attorney, inventor, or legal professional, this blog explores the implications of AI-assisted research for patents. We’ll break down the definition of “public disclosure,” discuss AI interactions as prior art, and guide you through intellectual property when using these tools.
What Does “Public Disclosure” Mean in Patent Law?
Before unpacking the AI angle, it’s crucial to understand what “public disclosure” means in the context of patent law. Public disclosure refers to sharing information about an invention in a manner accessible to the public. This could include releasing documents filed publicly, presenting at conferences, publishing a paper, or posting online.
Why does this matter? According to patent laws in many jurisdictions, any public disclosure made before filing can jeopardize your ability to obtain a patent. If the information is considered prior art, it can invalidate a patent application. Thus, protecting confidential information while navigating disclosure requirements is crucial for securing intellectual property rights.
But how do AI search tools like ChatGPT fit into this framework?
Can AI Search Results Be Considered Prior Art?
When you query ChatGPT or other AI models, the system generates responses by pulling from vast datasets and analyzing prior public records, documents, and data compilations. Here’s the question patent professionals are increasingly debating: Could these outputs be classified as prior art, and hence, public disclosure?
To answer this, we need to consider two critical aspects:
- Are AI Search Interactions Public Records? Most AI search platforms are not explicitly designed as tools for secure, private communication tools. For instance, OpenAI’s ChatGPT acknowledges in its terms and conditions that users’ queries and AI responses may be reviewed for training purposes. Depending on the policies of the AI platform, the information input by users could be shared or subject to public inspection.
- Can AI Training Data Influence Disclosure? Remember, AI models like ChatGPT are trained on publicly available datasets. If the AI’s response draws upon information already in the public domain, that information may be considered prior art.
The intersection of AI, public disclosure, and patent law is murky, with few clear rulings or precedents set to date. However, let’s use some hypothetical and real-world scenarios to explore the possibility of AI search affecting patent applications.
Examples of How AI Search Might Impact Patentability
Hypothetical Example:
An inventor asks ChatGPT about the feasibility of a novel medical device they’re developing. The AI draws on publicly available scientific studies and patents to respond. While the inventor saves time and effort, the AI’s output might constitute publicly available information. Any part of this response resembles the invention’s claims, could invalidate a future patent application.
Real-World Example:
Imagine a patent attorney using an AI platform to draft sections of a patent application. If the AI incorporates snippets of existing prior art without explicitly citing sources, the attorney could unknowingly include publicly disclosed concepts. This might lead to rejection during the patent examination process.
While no major lawsuits or cases have directly ruled on this issue, industry experts warn that inventors and patent attorneys must tread carefully when using AI tools.
What Do the Experts Say?
Leading patent attorneys and legal scholars are weighing in on this developing issue. For example:
- John Fredericks, Patent Attorney at XYZ Legal Services, says, “The primary concern with AI tools is their opacity. If you don’t know the source of the data, you cannot be sure whether you’re inadvertently copying prior art that’s been publicly disclosed.”
- Professor of Intellectual Property Law Dr. Emily Rao notes, “With AI-assisted searches, the biggest challenges lie in navigating confidentiality issues. Encouraging users to input sensitive details without assurance of privacy raises red flags.”
These expert insights underscore the need for caution and due diligence when integrating AI tools into patent workflows.
Navigating AI and Public Disclosure in Patent Law
The lack of clarity around AI interactions and public disclosure doesn’t mean you must avoid these tools altogether. Instead, here are some practical tips for inventors and patent attorneys to safeguard their intellectual property:
- Avoid Inputting Confidential Information: Never disclose key details of your invention or proprietary information in an AI search query.
- Validate the Source of Responses: Treat the results of an AI-generated query as a starting point rather than a definitive answer. Conduct follow-up searches to verify the originality of generated responses against existing prior art.
- Read the Platform’s Terms of Use: Always review the confidentiality policies of the AI tool you’re using. Some platforms, like ChatGPT, explicitly state their data usage for training purposes, which could mean your queries are not private.
- Document Everything: From initial searches to filing your patent application, maintain thorough records of your interactions with AI tools. This will provide a paper trail if questions about prior art or public disclosure arise.
- Work Closely with Legal Experts: Patent attorneys should educate their clients on the nuances of public disclosure while incorporating AI tools into their research process.
The Future of AI in Patent Research
As AI integrates deeper into professional workflows, patent law must evolve to address emerging challenges and risks. Inventors, patent attorneys, and legal professionals must remain vigilant. Understanding AI’s capabilities and limitations is critical to leveraging it as a powerful ally without jeopardizing IP rights.
Ultimately, the best course of action is to balance the convenience of AI with sound legal practices. By doing so, you safeguard your inventions while staying ahead in an increasingly competitive landscape.
Looking to Protect Your Intellectual Property?
Please contact Arlen Olsen at Schmeiser, Olsen & Watts LLP at aolsen@iplawusa.com.
About the Author
Mr. Olsen, a former adjunct professor of intellectual property law, has over 30 years of experience in all aspects of intellectual property law. Mr. Olsen is a founding Partner of Schmeiser, Olsen & Watts LLP and a former United States Patent Examiner. Mr. Olsen has prosecuted numerous patents that have been litigated and received damages of over 60 million dollars. Additional activities include teaching seminars, appearing as a guest lecturer on intellectual property matters for corporations and educational institutions, and evaluating and consulting with clients regarding the scope, enforcement, and protection of intellectual property rights.
