(123rf.com image)

Employee theft of trade secrets is not only on the rise but is now easier than ever before due to widespread access to rapidly evolving and affordable technology. Employees continue to use personal email accounts, thumb drives, phone cameras, personal cloud-based accounts (such as OneDrive, Google Drive, and iCloud), and external hard drives to mass-transfer, store, and retain their employer’s trade secret information post-separation. But are these types of technological advances the biggest threat to trade secret protection?

With the rise in popularity and availability of generative artificial intelligence (“gen AI”), many employers fear that safeguarding their trade secret information in the future may become even more difficult.

How employees are using gen AI in the workplace

Employers may be surprised to hear that, according to one survey recently conducted by McKinsey & Company, 61 percent of workers are either using or plan to use Gen AI in conjunction with their work.  In some instances, employee use of AI is employer-driven, such as where employers automate information/knowledge management or customer service functions. But in many instances, employee use of AI is employee-driven, and the employer may even be unaware of its use. Also according to McKinsey & Company, at least three times more employees are using gen AI for a third or more of their work than their leaders imagine, with millennials 1.4 times more likely to report extensive familiarity with AI tools than peers in other age groups.

For example, employees are using chatbots such as ChatGPT to draft documents and other content for work purposes, with some even inputting their employer’s confidential code and asking the chatbot to fix errors. Otter AI Chat is another popular chatbot used by employees to obtain real-time automated transcription and summarization of workplace meetings. Not only do these chatbots retain and store the information input, in many instances they use the information to “learn” and generate future output that they provide to other unrelated users.  

As of early April 2025, ChatGPT reported having more than 800 million users, with more ChatGPT users (upwards of 67 million) located in the United States than in any other country. At present Otter AI reports having upwards of 25 million users worldwide, with the tool having transcribed more than 1 billion meetings.   

How Gen AI could threaten trade secret protection

An employee’s public disclosure of a trade secret — including via gen AI — could, potentially, destroy the “secret” nature of the item, stripping it of protection under the federal Defend Trade Secrets Act (“DTSA”) and other very similar and widely-adopted state laws. This is because, for information to receive protection under these laws, a trade secret’s owner must take reasonable measures to keep the information secret, and must prove that the information derives independent economic value from not being generally known or readily ascertainable through proper means by someone who can obtain economic value from the disclosure or use of the information.  

Widespread or even inadvertent publication on gen AI platforms will likely prevent a company from meeting this bar. Indeed, some courts have observed that, once a trade secret is publicly disclosed via publication on a publicly accessible platform, its owner will be unable to avail themselves of the DTSA’s protections.

How employers can protect against theft of trade secrets via gen AI

Standing in stark contrast to the rapid development and popularity in employee use of gen AI in the workplace is how slowly companies are recognizing their associated risk, let alone developing a risk mitigation plan. As noted above, employers are obligated to take reasonable measures to protect their trade secrets from public disclosure, else the information can lose its trade secret status. This means that employers must stay abreast of the rapidly changing gen AI tools that their employees may potentially use to publicly disclose their trade secrets. Further, because many employers are unaware that their employees are even using gen AI tools in conjunction with their work, employers often miss that a departing employee could retain confidential information in various personal chatbot and other gen AI accounts post-separation.  

Failure to demand and ensure the return and/or destruction of any such information may also undermine its trade secret status and expose its owner to an argument that it failed to take required reasonable measures.

Employers should regularly and openly communicate their expectations to employees regarding the permissible and impermissible use of gen AI in conjunction with their work, and explain the risks presented to the company and to the employee individually if they fail to comply. Many employers elect to include these provisions in employment or confidentiality agreements. Employers should also carefully draft and implement internal policies governing the permissible use of AI. These policies should specify what gen AI services employees may and may not use in conjunction with work, with greater — if not outright — restrictions on employee use of publicly available systems that cannot or do not distinguish between confidential and non-confidential input.  

Employers may also elect to implement automated IT features that prohibit employees from uploading more than a certain volume of information to a gen AI platform using company-owned devices or servers, or that notify IT when an employee is frequently accessing or uploading data to such a platform. Employers must also regularly train employees on what constitutes confidential and trade secret information and, where possible, include clear labeling of the information as such, so that employees understand what they may and may not share with gen AI. Of course, some companies also elect to build and input information into their own gen AI app, and grant limited access to such applications on a need-to-know basis within the company.   

In sum, employers must continue to monitor changing technology and, accordingly, adapt the measures they take to protect their trade secrets from disclosure and preserve their secret nature. Provided employers do so, they will retain strong arguments that they took reasonable measures to protect their trade secret information, despite the rise in employee use — and potential abuse — of gen AI.

Jennifer S. Cluverius is a shareholder with Maynard Nexsen.