Legal Tech • May 2025

New Risks, New Opportunities

Protecting trade secrets in the age of AI.

Written by Jillian Beck

The capabilities of artificial intelligence (AI) continue to expand every day, providing new and more efficient ways to work. But with these opportunities come a whole new set of concerns. Companies whose employees utilize AI face the serious risk that their most valuable confidential information will be compromised. In the face of these new challenges, attorneys play a critical role in helping clients develop robust policies and practices to keep their “crown jewels” safe.

Virtually all businesses have “trade secrets,” meaning business information that has commercial value because it remains secret. The most famous example of a trade secret is the formula for Coca-Cola. However, trade secrets can also be much more mundane, like customer lists, financial information, or collections of data.

AI tools, such as ChatGPT, are a powerful resource for employees looking to perform their work more efficiently. ChatGPT is a generative AI model that accesses large amounts of data to generate responses to user queries. Employees can use ChatGPT for a broad variety of tasks, including drafting documents, summarizing information, or debugging software code.

To qualify for protection of trade secrets, employers must take reasonable steps to protect their secrecy. But information entered into generative AI tools is not always secure. For example, ChatGPT saves information that users input, and may use that information later to generate responses to other users. As a result, employees could unknowingly compromise trade secrets by inputting confidential information into AI tools.

Companies must work diligently to ensure that they are doing everything possible to protect their trade secrets in this new environment. The following strategies can help guard against unwanted disclosures:

Identify trade secrets: The first step in a robust trade secret protection plan is for companies to understand what their trade secrets are. If in doubt, valuable information should be protected as if it were a trade secret.

Limit access to confidential information: Trade secret information should only be accessed by employees who need that information to do their jobs. Companies should utilize password protections to ensure that trade secrets are shared on a need-to- know basis.

Establish policies for the use of AI: Company policies should make employees aware of the procedures in place to protect trade secrets, including clear guidelines for the use of AI tools. For example, companies should advise employees never to input confidential information into AI tools, or to obscure or remove identifying details for certain types of data. These policies should be regularly reviewed to make sure they remain up to date as AI continues to evolve.

Implement robust training programs for employees: Of course, written policies will not provide any protection at all unless employees follow them closely. Employees should receive regular training on the importance of protecting trade secrets and the potential risks of inputting information into AI tools.

While AI presents significant new challenges for protecting trade secrets, the above strategies can help ensure that companies can leverage these new tools while maintaining their legal rights.

This article, which was originally published in the October 2024 issue of Circuits, has been edited and reprinted with permission. Circuits is the e-journal of the State Bar of Texas Computer & Technology Section. More information can be found at sbot.org.

JILLIAN BECK is a partner in Prichard Oliver Montpas, a trial and appellate firm in San Antonio. Her practice focuses on complex commercial litigation, often involving trade secrets and technology issues.