Nora Roberts Foundation Generative AI Policy

The Nora Roberts Foundation stands against the use of Generative AI and LLMs to create content, and as a replacement for human teachers, therapists, and companions. We oppose this use for ethical, environmental, and humanitarian reasons.

All staff and board of the Nora Roberts Foundation agree to this policy and will not utilize Generative AI to create any Nora Roberts Foundation content. This includes, but is not limited to, emails, websites, reports, social media, and graphics.

The Nora Roberts Foundation will not knowingly grant to any organization who teaches or uses Generative AI to make creative content. This includes teaching students how to create essays, stories, visual art, videos, music, or computer programs with Generative AI, using Generative AI to create published art or written works, and using chatbots to provide charitable services such as therapy or tutoring or other conversation beyond a simple search or password reset.

We include in Generative AI, creative content, writings, images, music, video, or code made by prompting a program such as ChatGPT, OpenAI, Sora, DALL-E, Midjourney, Claude, Gemini, GitHub Copilot, Suno, Sora, Stable Diffusion, Grammarly AI tools, and more. Generative AI and LLMs create works based on their consumption of training data, acquired by consuming the work of countless authors, artists, and internet users.  Without the express consent of the original author/artist/creator, we view the resulting output to be an infringement of intellectual property and a violation of copyright. We believe creators must explicitly consent to use of their content for the Generative AI’s models sourcing of content to reach ethical standards.

Regardless of the ethical implications, the energy needed to power the data centers used to train and run LLMs and Generative AI models impacts society. And the increased energy use, water consumption, and pollution affects people who live near data centers. While some AI is used for the good of humanity, such as in science and medicine, we don’t believe use of Generative AI for content creation reaches a level of benefit to humanity to warrant the environmental cost.

We also consider the negative effects of Generative AI content on humanity. Research by MIT indicates that using ChatGPT to write essays contributes to cognitive decline in students, and that the more users talked to ChatGPT, the lonelier they felt. Chatbots played a significant role in the deaths by suicide of multiple children and adults. In other cases, chatbots engaged in sexual conversations with minors, and there are deepfake videos online of pornography using nonconsensual images of women and children.

As AI use in general increases, it becomes harder to avoid. We do see a difference in Generative AI for content creation and other AI tools, such as simple translators, editing software, and simple grammar and spelling checkers that point out common errors within the user’s own created content. AI research tools are also practically unavoidable. We believe in an abundance of caution when using AI for research or translation, as Generative AI is known to “hallucinate” and can return false information. AI translators are not sophisticated enough to serve as a replacement for a human translator. However, we do not include this type of AI use in prohibited activities for staff/board members, and this type of AI use does not impact our decisions regarding potential grantees. Research on AI and its impact may require its use and is also not a prohibited activity.

We do not condone the use of AI writing tools that generate written work or rewrite a user’s writing into a different style. If we suspect an application or a scholarship essay has been AI generated, we will run it through a Generative AI detector. We will not base decisions solely on the result of the detector, as we are aware that false positives are possible, but it will factor into our decision-making process.

We publish our Generative AI Policy to provide transparency to potential and current grantees about our operations and priorities. If there are questions about the Nora Roberts Foundation Generative AI Policy, grantees and applicants can contact kat@norarobertsfoundation.org and adanna@norarobertsfoundation.org.

AI Definitions

“Artificial Intelligence or AI for short is generally defined as the general ability of computers to emulate human thought and perform tasks in real-world environments – such as perceiving, analyzing, understanding and collating for synthesizing.“
(https://systemicalternatives.org/wp-content/uploads/2023/06/rs-09-ai-1.pdf pg. 4)

Generative AI: “Generative AI refers to deep-learning models that can generate high-quality text, images, and other content based on the data they were trained on.”
(https://research.ibm.com/blog/what-is-generative-AI)