When it comes to thinking through how AI-generated content may affect your company, rather than focusing on how your company can use it, should your company be preparing for how to protect itself from its use?
Here is a general guide for a few discussion points that may be applicable to your company:
For businesses, do you have internal policies in place for using these tools and/or limiting their use?
Are your employees and/or independent contractors restricted from using AI tools, either through web restrictions and/or company policies on what content they are allowed to input in these tools. For instance, if an employee were to use an AI program to record notes during a meeting, would this type of use require advanced approval from their manager or supervisor? Are policies in place at your company to prevent this type of input going into these AI-content generators?
See here for more on how employees and executives can be inappropriately sharing information in AI models. It also appears that large companies are also taking notice, by implementing corporate guidance on use of ChatGPT and similar programs.
For companies sharing content from AI-models, do you have necessary provisions in your agreements for your employees, independent contractors, and/or your users of this content?
Are your agreements sufficient to protect your company if employees or independent contractors use any AI-generated output in your products and services? Your company may be unknowingly violating others’ intellectual property rights, breaching license agreements and/or other agreements, and you may not want to find that out down the road when you receive a take-down notice or a demand letter.
If using an AI model or modified AI model, is your company complying with license requirements?
If your company is using and/or relying upon an AI model that is released with an open source license, a threshold step is to identify what the terms and conditions of the AI model’s owner permit and/or require for using their model. If your company is using these AI models, has your company confirmed its compliance with those licenses (such as following requisite conditions for the usage, distribution, or redistribution of those AI models)? Depending on the end user, some licenses may require specific notices for altering the AI model and providing that AI model for others to use. In addition, some licenses may require the users of the AI model or derivatives thereof to be restricted in how the model can be used. These are more specific examples of issues that may need to be addressed, but one can see how quickly adopting an AI model, making changes, and pumping it out to end users without adequate safeguards could expose your company to significant liability.
Thinking through how to incorporate AI-models into your company can be a fun, initial step for your company. While doing so, however, it may be useful for your company to also consider how these tools can be used to expose your company to additional risk and future liabilities.
If you want to speak to a tech and IP attorney or have other tech-related legal questions, you can contact me here.