For universities and other research organisations, protecting innovations through IP (intellectual property) has never been simple. For example, patent rules, which in most countries force inventors to file a patent before the innovation becomes public, can conflict with the desire to publish and share research freely with the world. But without a patent, innovations may be difficult or impossible to commercialize.
Now AI has added a new layer of complexity to this question. The capabilities of generative AI have raised all kinds of new issues – can AI be an inventor? How should AI-assisted inventions be judged? And just what is patentable with AI anyway?
Let’s start with the first question: can AI be an inventor?
The short answer is no, at least not yet. Here's what courts around the world are saying:
The German case is particularly interesting. Stephen Thaler, who applied for patents for inventions generated by his AI system DABUS, was ultimately allowed to be named as the inventor, despite claiming no creative input. This paradoxical judgment highlights the complex nature of AI-assisted inventions.
AI-Generated Ideas: A Double-Edged Sword
AI's ability to generate numerous ideas raises two critical questions:
The USPTO recently sought public commentary on this issue, underlining its significance in the patent landscape. No government patent office has actually answered these questions anywhere in the world, as far as I know - although all of them are grappling with these and other issues around AI.
USPTO Guidance on AI-Assisted Inventions
The USPTO recently issued guidance on AI as a tool for creating new innovations, or AI-assisted inventions. This guidance emphasizes that patentability of AI-assisted inventions will be assessed case-by-case. Applicants must clearly highlight the human inventor's role. But given the way the guidance was written, questions remain about what constitutes a significant contribution by a human inventor.
However, with the Supreme Court's recent decision to strike down "Chevron deference," the future application of this guidance remains uncertain.
Key Takeaways for Innovators
Given the current uncertainty and fluid environment involving decisions around AI-assisted inventions – and AI patentability – here are some important points to remember:
✅Human-Centric Invention: Only humans can be patent inventors.
✅Detailed Documentation: Meticulously document any AI use in the innovation process. This is crucial for patent applications and potential future litigation.
✅Clear Attribution: Clearly define and document the role of each human inventor and non-inventor in the innovation process.
And for our friends in the world of Tech Transfer specifically, here are our suggestions:
✅Develop detailed guidelines for researchers so that they know what to disclose about AI.
✅Create simple checklists or forms about the use of AI in research to capture necessary information for protecting your IP.
✅Maintain open communication with researchers about their innovation processes. Remember - a quick face to face discussion is worth a thousand emails.
What are your thoughts on AI's role in innovation? How is your organization adapting to these changes? Let's continue this important conversation in the comments below.
And if you want to learn more about AI, IP and tech transfer – join our course this fall with AUTM and Peter Bittner. It’s called “The Future of Tech Transfer: Leveraging AI to Find & Secure Licensing Deals, Research & Find Your Best Company Licensee”.
It runs from October 15 – November 7.
The goal is to provide you with the necessary information to integrate AI into your own process – and to create your own roadmap for doing so.
Places are limited so sign up now!