The UW Libraries seeks to advance research for the public good by ensuring everyone has access to research produced by our world class institution. We think critically about how our scholarship is produced and shared. As artificial intelligence (AI) becomes more mainstream, we hold closely these values in choosing how our scholarship is used in the production of AI tools and who we partner with on future projects.
July 2024 edition
Introduction
UW Libraries has long-standing values that can accommodate our potential work with artificial intelligence (AI). This framework guides UW Libraries staff in applying our values to operations and decisions related to AI.
We value
User-Centeredness
Members of our community interact with and may benefit from AI tools, and we value a user-centered approach to AI technologies.
We do this by:
- Prioritizing transparency in our communications about the integration of AI tools in Libraries services, collections, and spaces.
- Investigating and articulating ethical or privacy considerations raised by the AI tools we create, purchase, or support.
- Critically analyzing tools, projects, and outputs, and supporting our users in doing the same. This includes encouraging users to consider the context and potential biases of information provided by generative AI tools.
- Serving as a resource for users navigating uncertainty in this complex and quickly changing landscape.
- Supporting the emerging need for materials that may be useful for AI research and advocating for license agreements that allow for Libraries’ resources to be used in research, teaching and learning, and creative endeavors.
Collaboration
We partner with a diverse set of colleagues inside and outside of the Libraries to engage with AI work on our campuses.
We do this by:
- Seeking input from stakeholders as we consider adopting AI tools to provide Libraries’ services.
- Sharing our decisions regarding tools, projects, and subscriptions related to AI.
- Engaging with our communities to understand their activities and concerns related to AI.
Equity
We acknowledge and address both the harms and opportunities that AI tools present in the context of diversity, equity, inclusion, and accessibility (DEIA).
We do this by:
- Helping users and ourselves understand how AI may perpetuate existing inequities, biases, and racism, and prioritizing tools and initiaitves that address these harms.
- Recognizing and championing uses of AI tools that benefit marginalized communities.
- Identifying and mitigating harms in potential Libraries AI projects, including tools that we license or create.
- Ensuring that AI tools licensed by UW Libraries meet UW-IT Accessibility Guidelines.
Creativity
We embrace a culture of exploration, experimentation, and reflection to improve services, anticipate needs, and manage change, and will embrace that same culture when it comes to artificial intelligence.
We do this by:
- Encouraging the exploration of AI-based tools and services.
- Leveraging AI for the benefit of our users and communities.
- Creating a culture of intentional experimentation around AI-based projects and solutions, and empowering staff to propose new ideas that address our shared challenges.
Sustainability
We engage in critical analysis and holistic decision-making when considering AI’s positive and negative impacts on our ability to steward the human, environmental, physical, digital, and financial resources with which we are entrusted.
We do this by:
- Actively weighing the benefits of AI tools and workflows against potential costs to Libraries staff, users, and the wider community.
- Investing appropriately in AI projects to ensure that they have sufficient technical resources and staffing to be successful.
- Supporting community-controlled AI infrastructure to ensure that these tools are a part of an open scholarly ecosystem.
- Prioritizing transparency around pricing and terms of use in contract negotiations for AI tools and services.
- Investing in ongoing training to support staff development in a rapidly changing environment.