Many employers are already dealing with employees using Artificial Intelligence or Generative Artificial Intelligence (collectively “AI”) in their day-to-day job duties, communications, or work product.

It is problematic that employees may be using (or trying to use) AI completely on their own, outside of employer approval, knowledge, guidelines, or policies. The best way for an employer to put some shape or boundaries on employee use of AI is to implement an Artificial Intelligence Policy and conduct one or more meetings, and regular updates, on how the employer is operating under the policy and which AI tools are and are not allowable for employee use.

While AI itself is not novel, the availability of AI tools has rapidly expanded so that today, almost anyone can access AI. Many AI tools can alter pictures, videos and voices, resulting in a potentially dangerous situation in the workplace. Almost any employee can jump on the web, access a free AI tool, and give it a shot – all without the employer’s knowledge. Without workplace standards or policies governing the use of AI, the employee might pass the AI work product off as their own, even though the foundation or premise of the work product may be unfounded or simply wrong. This scenario presents considerable risk for the employer – especially employers of experienced or trusted employees that are unlikely to question, much less fact check, the work product or communication.

AI tools, particularly GAI tools, digest huge amounts of literature or text from millions of sources – including potentially out-of-date sources – and then predict what a human most likely would say in response to questions. The accuracy of the responses is limited by how well the questions are crafted and the quality of the data the AI tool uses. To an uninformed questioner, AI may provide obvious mistakes or factual inaccuracies in response to a question. The potential result is employee work product, or coworker communications about products, which are inaccurate but appear impressive and honest to management, coworkers, or customers.

In addition to potential perils associated with reliance on AI output, employers should consider what confidential information an employee might put in the AI tool.  The use of AI tools is generally subject to terms and conditions that give the owner of the AI tool broad rights to use the data that that AI tool users input. Putting the company’s proprietary or confidential information, or any personal data, into an AI tool is a risky proposition.

To mitigate risk and stop or slow down renegade or stealth employee use of AI in the workplace, employers should implement an Artificial Intelligence Policy which, at least, requires an employee to obtain authorization from the employer on key components of the proposed use, such as,

1. Identification of the proposed AI model or tool to be used and its provider;
2. A description of the project for which the AI tool will be used and a detailed explanation of the proposed use, including whether company protected information might be used or disclosed in the project;
3. An assessment of the accuracy of the proposed tool to be used; and
4. Any additional information that will be useful in understanding the request for authorization.

Companies should also identify the criteria they will use to determine whether to authorize AI or GAI use, and the post-authorization conditions for employee use. The policy should also include penalties for unauthorized use.

Bodman cannot respond to your questions or receive information from you without establishing an attorney-client relationship and clearing potential conflicts with other clients.  Thank you for your patience and understanding.