Intellectual Property Considerations
One advantage of ChatGPT is how the platform can streamline the generation of content. ChatGPT allows a user to pose a question, which OpenAI calls the “Input,” and in return, receive an “Output” response generated from ChatGPT’s machine learning algorithm pulling an answer from ChatGPT’s available sources. The Input and Output are defined together as “Content.” As set forth in the Terms, the user owns all of the Input and Output. However, the Terms restricts users from using the ChaGPT in any manner that infringes upon, misappropriates, or violates a third party’s rights. This leaves room for third parties to have a claim infringement or misuse in parts of the Content. As such, companies seeking to leverage ChatGPT Content in their business should assess the intellectual property risk with counsel.
Likewise, OpenAI puts shifts privacy compliance obligations to the users leveraging the Content. The Terms require users to represent and warrant that if users are processing personal data, they are complying with all applicable data privacy and data protection laws. Additionally, it is important for users to understand that they are granting OpenAI a license to utilize the Content to improve their services. However, users can opt out from their data being used by OpenAI by filling out an opt-out form.
Commercial Use of Content Considerations
The Terms expressly allow for a user to utilize the Content for commercial purposes. However, there are legal risks in doing so. The Terms shift the liability relating to the Content to the user and make it the user’s responsibility to ensure the Content is not infringing third-party rights and applicable laws. For third-party rights, a proper intellectual property assessment should be completed to ensure that there are no third-party rights to the Content. For applicable laws, there should be a proper privacy assessment to confirm that no applicable privacy laws are violated by the use of the Content.
The Terms also notify the user that the Output is not guaranteed to be accurate information and places the onus on the user to verify and confirm that the Output is accurate information. Thus, relying upon the Output without due diligence in a commercial manner could lead to negligence claims and other claims related to false information being used for various commercial purposes.
The Terms restrict users in a unique way. A user cannot represent that the Content was human-generated when it is not. Such representation is considered a breach of the Terms and would result in potential legal liability.
As the adoption of AI in everyday business use continues, companies may have to assess the legal considerations before implementing the technology in their business models. ChatGPT is only one specific example. However, the analysis of OpenAI’s Terms highlights the potential risk of leveraging generative AI-enabled solutions.
For more information, please contact any attorney with Frost Brown Todd’s Artificial Intelligence team.