Articles | Jan 18, 2024

AI Demystified: Decoding Generative AI Platforms

Rochester Business Journal

Read the Article

Complying With Generative AI “Terms of Use”

In 2023, generative artificial intelligence (AI) platforms experienced significant growth, signaling a transformative era in how we work and live. These platforms are poised to reform various aspects of life through the use of applications that deploy technologies such as natural language processing and the synthesis of images.

People across various industries, professions and experience levels are using generative AI to improve efficiency or merely out of curiosity. Therefore, it is not surprising to learn that managers, executives and business owners feel pressure to jump on the AI bandwagon and implement plans that incorporate the use of AI to improve productivity. While certain benefits of generative AI are apparent, there could be harmful results if generative AI is introduced in a business setting without first reading the fine print.

It seems that this transformative technology has caused many people to overlook an obvious pitfall. These generative AI platforms are
being provided by a third party, often as a software as a service (SaaS) offering. Accordingly, like any other third-party service offering, use of a generative AI platform is governed by a third party’s end user agreement and other third-party terms and conditions. As users engage with these platforms (especially those available for free), it is crucial to review, understand and comply with the terms and conditions, often referred to as “terms of use,” that accompany and govern the AI platform. This article sheds light on certain key aspects of generative AI platforms’ terms of use, outlining things users should consider before diving into the world of generative AI.

User Input: Granting Permissions and Intellectual Property

What can a generative AI platform do with the data a user inputs? While people ponder this question as if it was shrouded in mystery, the answer is to review the terms of use, at least as a starting point. Generative AI platforms often require users to grant certain permissions to the third-party provider for the data inputted into the generative AI platform by the user. For example, by inputting data into the generative AI platform, the user may grant the third-party provider the right to use their data for developing and improving the generative AI platform for all users. If the user does not want certain data to be used in this way, the user may have the ability to opt out. Ultimately, the terms of use will explain the user’s rights with respect to the input.

Generative AI Output: User’s Rights (Intellectual Property Considerations)

We are often asked about the manner and extent to which one can use the output from a generative AI platform. Users must carefully review the terms of use to determine the right to use output including ownership of the output (including intellectual property rights), permitted and prohibited uses, and attribution obligations. One platform may only grant the user a license to use, modify or commercialize the content created on their platform, while another platform may assign ownership of the output to the user. For businesses wishing to use such content, especially for commercial use or reproduction, it is critical to understand these terms and assess whether they align with their intentions for the generated content.

Data Privacy and Security

Generative AI platforms may collect and process user data to improve their algorithms or provide personalized experiences. Often terms of use will include a hyperlink to a company’s privacy policy, which should be read in conjunction with the terms of use. Users should carefully review the platform’s privacy policy to understand (1) the types of data collected (i.e., personally identifiable information) and (2) how data shared on the platform is maintained, stored and shared (i.e., with a company’s affiliates or third parties). Further, it is always advised that users avoid including personal or confidential information in any prompts submitted on the AI platforms. Other important sections of the privacy policy, include terms on encryption, data anonymization or aggregation, and cybersecurity standards — in other words, the measures in place to keep data safe and prevent security incidents. An understanding of key privacy policy provisions is helpful to instill confidence in the platform’s commitment to data protection.

Responsible Use and Ethical Considerations

As generative AI becomes more powerful, ethical considerations come to the forefront. Terms of use may explicitly obligate the responsible use by user and discourage the creation of harmful, offensive or misleading content. Some providers have a separate acceptable use policy to moderate content in violation of the terms of use. Such acceptable use policies often include prohibitions on content that discriminates against others, violates the rights of third parties, spreads misinformation or false information or is otherwise unlawful. Business owners must be aware of these guidelines to avoid potential consequences for violating the platform’s terms and should educate and train their employees, agents and contractors accordingly.

Disclaimers, Liability and Indemnification

Generative AI platforms typically include disclaimers regarding the accuracy and reliability of their services and stipulate that the services and platform are provided “as is” to its users. Users should understand the limitations of the platform and the extent to which the provider assumes liability for any potential issues, including errors or unintended consequences arising from the use of the generated content. Further, the terms of use will likely stipulate limitation on damages for claims arising related to use of the platform — for example, that the third-party provider will not be liable for damages of any kind and/or provide for a cap on damages of a nominal amount. Additionally, terms of use may specify one party’s responsibility to indemnify and hold the other harmless against certain claims or damages resulting from use of the service. Be mindful that the disclaimers, limitations of liability and indemnification provisions will vary from provider to provider and will likely be different for a free offering as compared to a paid offering. Free offerings are often provided “as is,” with little to no protection to the user, while paid offerings often provide more protection from a warranty and indemnification standpoint. One important area where this difference comes to light is in the context of copyright infringement with respect to the output. Users of a paid offering are often indemnified if output is infringing, but may bear the risk of infringement for use of a free offering.

Changes to Terms and Service

Generative AI platforms usually reserve the right to update their terms of use periodically as the third-party provider sees fit. Users should be vigilant in checking for updates and remain informed on how changes may affect their existing use of the platform. Platforms may provide notifications or require users to accept updated terms, emphasizing the importance of understanding any modifications. In certain industries, platforms may be required to disclose the terms of use on an annual basis. It is also to be expected that the terms of use state that a user’s continued use of the platform will constitute acceptance of such revised terms.

There is immense potential for generative AI platforms to lead to optimization and innovation in the work place, however, users must not forget that, at the end of the day, such platforms are just another product or service governed by a contract. Users must approach generative AI with a clear understanding of the terms of use. By carefully reviewing and comprehending such terms, users can navigate the ethical, legal and practical aspects of engaging with generative AI platforms. This knowledge empowers users to make informed decisions and use these powerful tools responsibly, making for a positive and safe experience in the evolving landscape of AI. The Phillips Lytle AI team is available to review or interpret agreements related to your utilization of an AI platform or to assist with other AI-related issues. Stay tuned for the next installment of this series as we continue to dive deeper into the emerging world of AI.

Richard J. Marinaccio is a partner at Phillips Lytle LLP and leader of the firm’s Technology Industry Team, Intellectual Property Team and chair of the firm’s Technology and Innovation Committee. He can be reached at rmarinaccio@phillipslytle.com or (716) 504-5760.

Dorothy E. Shuldman is an attorney at Phillips Lytle LLP and a member of the firm’s Corporate and Business Law Practice and Intellectual Property Team, focusing on trademark and copyright law. She can be reached at dshuldman@phillipslytle.com or (716) 504-5778.

Related Insights

View All