The U.S. Patent and Trademark Office (USPTO) has issued new guidance to address the growing role of artificial intelligence (AI) in patent law, highlighting the complexities and challenges of determining patent eligibility and inventorship for AI-assisted innovations.
Released in 2024, the AI Inventorship Guidance and AI Eligibility Guidance mark significant milestones in clarifying the rules for AI-driven innovation. These guidelines aim to provide clarity to inventors and patent examiners while navigating the legal landscape shaped by the U.S. Constitution and federal courts.
One of the central issues tackled in the USPTO’s guidance is whether inventions created with or by AI can be patented. According to U.S. law, patents must list at least one human inventor who has made a “significant contribution” to the invention. This follows the Federal Circuit’s 2022 ruling in Thaler v. Vidal, which held that only natural persons could be named as inventors on patents. The decision reinforced that AI systems, even when generating innovative concepts, cannot be listed as inventors under current law.
To address ambiguities, the USPTO introduced criteria to determine inventorship in cases involving AI assistance. A human’s role must go beyond routine contributions and involve substantial input into the invention’s conception. For instance, prompting or training an AI system for a specific solution may qualify as a significant contribution, but merely owning or overseeing the AI system does not.
USPTO Director Kathi Vidal emphasized that the guidance aims to embrace AI’s potential for innovation while ensuring adherence to legal standards. As long as a human’s input meets the significant contribution threshold, patents may still be granted for AI-assisted inventions.
In addition to inventorship, the USPTO addressed the eligibility of AI-related patents, particularly those involving abstract ideas. Under the Alice/Mayo framework, patents directed toward abstract concepts, such as mathematical formulas or general methods, must demonstrate an “inventive concept” or integrate the idea into a practical application to qualify for protection.
The AI Eligibility Guidance provides examples to illustrate these principles. For instance, a patent claim for a specific application of AI in hardware, such as a neural network integrated into a microprocessor, would likely be eligible. However, general methods of anomaly detection using AI are considered too abstract without a specific technological context.
Stakeholder feedback has been mixed. Some argue that the guidelines are overly restrictive, potentially stifling innovation by requiring extensive human involvement. Others contend that the rules may allow overly broad patents for AI-assisted innovations, leading to legal ambiguities.
While the USPTO’s guidance represents a step forward, it remains non-binding, meaning its principles could be challenged in federal courts. Legislative action may ultimately be necessary to resolve the ongoing uncertainties surrounding AI and intellectual property law.
Need Help?
If you have questions or concerns about AI guidelines, regulations and laws, don’t hesitate to reach out to BABL AI. Their Audit Experts can offer valuable insight, and ensure you’re informed and compliant.
