Find your IP experts

News

IP NEWS

AI and Inventorship: Best Practices for Securing Your Patent Rights

  Today, the world’s leading patent offices and courts – including the European Patent Office (EPO) and the U.S. and German courts – consider that an inventor must be a natural person. Consequently, an artificial intelligence system cannot be named as an inventor.

This issue is particularly significant in the United States. Naming the wrong inventor may result in a U.S. patent being held invalid.

The United States Patent and Trademark Office (USPTO) considers AI systems, including generative AI, to be tools, much like laboratory equipment, software, research databases, or any other tool used during the inventive process. While AI can generate ideas and produce data, it ultimately remains a tool that supports the human inventor.

Under U.S. case law, inventorship requires a person to have made a significant contribution to the conception of the invention. In this context, “conception” is defined as the formation in the inventor’s mind of “a definite and permanent idea” of the complete and operative invention. If a named inventor merely relies on AI-generated output without making an inventive contribution, their status as an inventor may be challenged. In patent litigation, this could ultimately result in the patent being held invalid.

Thus, now more than ever, the USPTO reaffirms the fundamentally human nature of inventions.

In that context, companies using AI throughout their innovation processes should consider maintaining records demonstrating sufficient human intellectual contribution should the issue ever arise.

Clearly Define the Technical Problem Before Using AI

The first step is to document the human origin of the inventive process.

Before using any AI tool, companies should clearly define the technical problem to be solved and the objectives of the project. Doing so helps demonstrate that the inventive concept originated from human reasoning rather than from the AI system itself.

Companies should also retain technical documentation, working notes, meeting minutes and internal communications demonstrating that the technical problem was identified and formulated by one or more human inventors.

Establish a Clear Framework for AI Use

AI should be used solely as a research and design support tool.

From an operational perspective, companies should favour internally deployed AI solutions or systems operating within secure, company-controlled environments. Where external AI platforms are used, particular attention should be paid to the applicable terms of use, confidentiality safeguards, and intellectual property provisions governing the generated data and content.

Furthermore, prompts submitted to AI systems should reflect prior technical analysis conducted by the R&D teams. Preserving a complete record of prompts and AI-generated responses can provide valuable evidence that AI merely assisted the inventive process rather than replacing the inventor’s intellectual contribution.

Document the Intellectual Choices Made by Inventors

Companies should retain records of all solutions proposed by AI, including those ultimately rejected.

The technical reasons why inventors selected certain solutions over others should also be documented. Meeting minutes, technical discussions among team members, and records of the reasoning that led to the final technical solution may all serve as valuable supporting evidence.

Likewise, any modifications, improvements or adaptations made by the inventors to the AI-generated proposals should be carefully documented.

Clearly Identify Human Contributors

Finally, companies should accurately identify every individual who contributed to the various aspects of the invention. This documentation may include a detailed allocation of each team member’s contribution.

Such traceability not only facilitates the correct identification of inventors when filing a patent application, but also provides valuable evidence should the status of an inventor be contested at a later date.

Conclusion

The use of artificial intelligence is fully compatible with obtaining patent protection. However, in a context where only natural persons may be recognised as inventors, it has become essential to demonstrate that the claimed invention reflects a sufficient human intellectual contribution—particularly in the United States.

The methodical preservation of evidence relating to the definition of the technical problem, the use of AI, the decisions made by the inventors and the development of the technical solution is now an essential safeguard for securing future patent rights.

Back to the articles