Rethinking IP Strategy in an AI-Driven World

By Norm Freda

For decades, patents have served as a cornerstone of innovation protection. A patent gives its owner exclusive rights to prevent others from making or selling an invention for a limited period of time in exchange for publicly disclosing the invention. This enables others to learn from, build upon, and improve the disclosed invention, effectively fueling the next wave of innovation, while rewarding the inventor for their creativity, ingenuity, and disclosure. However, some have questioned whether this bargain still works in today’s artificial intelligence (AI) landscape.

Billionaire investor Mark Cuban recently suggested that, with the rapid rise of AI, companies may think twice before filing a patent application. His reasoning is straightforward: once a patent application is published, it becomes public information. In most cases, US patent applications are eventually published and become available to the public. The published patent application can then be scraped, analyzed, and incorporated into AI databases.

That means competitors, or anyone in the world, can utilize AI tools to quickly understand, improve upon, or design around your invention – leveraging insights mined from the patent disclosures without the traditional investment of time, money, or manpower. In other words, your own patent application may be training the very AI tools that others use to compete with you in the marketplace.

Cuban predicts that some companies may increasingly turn to trade secrets rather than patents to protect some of their intellectual property. A trade secret is a piece of proprietary information that is kept strictly confidential to obtain and/or maintain a competitive advantage. The formula for Coca-Cola, for example, is most assuredly kept as a trade secret. Trade secrets, however, do not provide an enforceable limited monopoly like a patent and can be lost in the event of a leak, independently discovered, or copied/imitated.

Interestingly, the number of patent applications filed in the United States in 2025 fell by around 9% from 2024. This is a noticeable shift after years of steady growth and a record high number of US patent application filings in 2024. Increased costs and uncertainty in the economic and political environment have certainly played a large role in the recent decline in US patent application filings. Yet one cannot help but wonder whether companies reevaluating their patent filing strategies and keeping more trade secrets to better compete in the AI landscape, as Cuban suggested, was another key contributor to the decline.

Cuban’s comments highlight how technology, particularly AI, is constantly reshaping long-standing business practices and strategies including the protection of intellectual property. Regardless of whether his prediction comes to fruition, it challenges companies and inventors to reevaluate their strategy for protecting their intellectual property and has sparked discussion about the future of intellectual property in today’s AI landscape.

Norman K. Freda is an attorney at Fishman Stewart and practices in the fields of patent, trademark, copyright law, and litigation. He holds a BS degree in Applied Engineering Sciences from Michigan State University. Check out his bio.