Key Takeaways from "Generative AI and Ethics, Copyright, Privacy, and Security" (10/20/23)
Generative AI tools (and machine learning in general) can be biased in a variety of ways, and bias can happen at every stage of development. One way bias occurs is through datasets that misrepresent, exclude, or marginalize certain identities, communities, and practices. When models are trained on these datasets, they may reflect and amplify social prejudices and stereotypes represented in the dataset, or may fail to generate good outputs for underrepresented groups.
Generative AI may also provide factually inaccurate outputs, generate "fake citations," or misrepresent information in other sources. Gen AI tools can also be used to create manipulated and faked text, video, images, and audio. Safety guardrails and other automated practices cannot fully prevent this. Practicing critical information literacy and verifying information is especially important.
Copyright issues are currently being debated on two fronts: the legality of data scraping practices for generative AI that include copyrighted material, and whether works created with generative AI tools can be copyrighted. The US Copyright Office is undertaking a study of legal and policy issues related to generative AI, including an invitation for public comment.
When considering using generative AI in your work,
- be intentional - check to see if the journals/organizations/etc. you wish to share your work with have generative AI content policies
- be transparent - disclose your use of generative AI
- be specific - keep notes on your prompts and process for yourself and others
You can ask some or all of the following questions:
- Why do I want to use gen AI for this task, process, or project?
- What function specifically do I want this tool to perform? Ex: outline, critique, brainstorm, explore, refine
- How well does this use align with my personal values?
- What do I expect the benefits to be?
- What might be some costs or harms involved, to myself or others?
- How might I mitigate these costs or harms?
- How am I protecting the privacy and intellectual work of myself and others?
- Am I comfortable with my opinions or approach being influenced by gen AI?
- How might I verify information I am getting from gen AI?
It can also help to become more familiar with the datasets, development, and design goals of these tools in order to understand whether they will meet your needs.
Additional resources about these topics are available at: https://libguides.amherst.edu/genAI