Generative AI has made headlines in recent months after the large language model ChatGPT launched for free public use. ChatGPT set the record for the fastest growing user base, obtaining 100 million monthly active users in only two months.[1] By March, that number had grown to over 1 billion users.[2] The allure of ChatGPT is that it can quickly provide answers and write drafts, allowing users to get through their work faster and easier.
Its use in the workplace isn’t limited to just a few. Indeed, a recent survey by Fishbowl found that 43% of the responding professionals had used AI tools for work related tasks, including ChatGPT.[3] Perhaps even more significant, almost 70% of those who had done so did not disclose their use to their employer. Those employees work for companies big and small and across business sectors – from Amazon, to Nike, to Bank of America, to Twitter, to McKinsey.
These statistics highlight a key risk now facing every business that believes it has trade secrets worth protecting: whether or not your business has engaged with, or even considered, the potential implications of generative AI, you may be risking your trade secrets without even knowing it. Employees may be putting confidential business information into ChatGPT as prompts to do their work. For example, an employee may enter information about confidential product development strategies into ChatGPT to write a draft press release or marketing campaign materials. Or an employee may input portions of confidential computer code in seeking to solve a programming problem or to write a technical bulletin.
This input of company information into a third-party generative AI platform arguably could undercut the ability to claim the information constitutes a trade secret. Under Washington law, a key requirement to establish a trade secret is that the information be “the subject of efforts that are reasonable under the circumstances to maintain its secrecy.”[4] Trade secrets must be secrets, and the dissemination without confidentiality protections of information defeats a claim of secrecy. If organizations don’t have policies in place prohibiting employee use of ChatGPT, or clear guidelines restricting the use of confidential business information in connection with generative AI tools, a competitor might argue that the company failed to take reasonable efforts to keep its information secret. Given the growing prevalence of employees using generative AI without an employer’s knowledge, having internal controls in place to tightly limit the dissemination of trade secrets within the company may be more important than ever.
In our next article, we will discuss how the use of generative AI may impact an organization’s ability to claim their information is sufficiently “novel” to constitute a protectable trade secret.
This is the first article in our series “Mitigating Legal Risk in Today’s Business Climate – a Litigator’s Lens.”
[1] https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/
[2] https://www.demandsage.com/chatgpt-statistics/#:~:text=ChatGPT%20currently%20has%201.16%20billion,million%20users%20within%20two%20months
[3] https://www.fishbowlapp.com/insights/70-percent-of-workers-using-chatgpt-at-work-are-not-telling-their-boss/
[4] RCW 19.108.010(4)(b).
Brandi B. Balanda
Brandi has wide-ranging experience successfully litigating large, complex cases, from defending a senior executive against the largest trade secret claim asserted in Washington to prosecuting fraud, contract, and fiduciary duty claims related to the financing of major real estate development projects.