Polish Patent Office: AI Assistance Could Derail Patent Applications

Aneta Stuleblak, deputy president of the Polish Patent Office (UPRP), has issued a strong warning against using publicly available AI language models, such as ChatGPT, to assist in preparing patent documentation. She described such actions as "extremely irresponsible" and stated that it could jeopardize an entrepreneur's chances of obtaining a patent.

Stuleblak explained that information entered into these AI models is stored in their memory and can be subsequently used by other users. This poses a significant risk for inventors seeking to protect their intellectual property. The URPP is treating the issue of AI in industrial property as a major challenge and is focusing on educating the public about the potential pitfalls.

The deputy president's warning comes in the wake of author Olga Tokarczuk's admission to using AI in her writing process, which sparked debates about copyright for AI-assisted creations. The URPP is emphasizing that technical solutions created by algorithms, rather than humans, face significant hurdles in obtaining protection.

The office aims to increase the number of Polish patents and is actively engaging in educational efforts to discourage the uncritical use of AI tools. Stuleblak stressed that the URPP's goal is to foster innovation while ensuring the integrity and security of the patent application process.

“

Introducing publicly available language models, such as ChatGPT, with a description of a technical solution, in order for AI to help an entrepreneur prepare application documentation, is an extremely irresponsible behavior.