Top of page

Skip to main content
Library Information Digital Strategy

Artificial Intelligence at the Library of Congress

The Library of Congress has been researching and experimenting with machine learning (ML) and artificial intelligence (AI) since 2018. As these technologies transform how we access and share information at an ever-accelerating pace, the Library is taking a strategic approach to adapt and engage ML and AI opportunities ethically, responsibly and in support of our mission to Congress and the American people.

The Library embraces new technologies with care. With a mission to “engage, inspire, and inform Congress and the American people with a universal and enduring source of knowledge and creativity,” the Library preserves and provides access to a rich, diverse and enduring source of knowledge to inform, inspire and engage in intellectual and creative endeavors. Our roles as the research arm of Congress, home of the U.S. Copyright Office, and stewards of vast collections of the nation’s cultural heritage materials necessitate a managed and measured approach when adopting new technologies.

The Library's approach to AI builds on decades of technological innovation. Although AI presents new challenges and opportunities, the Library has stood at the forefront of innovation in knowledge and learning from its earliest days. We have spent decades maturing our technology governance structures, exploring advanced computing practices and building our capacity for data-driven decision-making. Today, we remain committed to providing authentic information and trustworthy services and making the best use of resources while respecting the rights of creators and users. The Library’s principles, use cases and experiments in AI are geared toward making these technologies work for our mission to Congress and all Americans.

The Library balances human judgment with machine capabilities. As AI technologies quickly evolve, human involvement and expertise are fundamental to the Library’s approach with these tools. Library staff, working with Congress, creators, users and peers around the world, are identifying and acting on the unique promises of AI while examining the potential risks of these technologies. Our human-centered approaches balance responsibility and structure with curiosity and flexibility.

Definition and Principles

In 2024, the Library developed a definition and set of principles to guide our work with AI tools and methodologies at the Library.

The Library’s Definition of AI

Artificial Intelligence (AI) refers to a machine-based system or combination of IT systems that can, for a given set of human-defined objectives, employ data, models or other technologies to create output based on predictions. In most AI systems, the factors used to form these predictions are not directly created or reviewed by humans but are generated through large numbers of calculations created within an AI model based on the data it is trained on. Depending on the objective for which they were designed, AI outputs may either be delivered directly to people or used in additional computation. The Library will generally use more precise terms and definitions to inform policy, technology, or practical decisions.

Principles for AI Use at the Library

Strengthen Trust: The Library will ensure we are collecting, stewarding and providing authentic and authoritative information in keeping with the mission of the Library.

Enhance Public Benefit: The Library will use AI technologies to support the broad availability and accessibility of our resources and services when, where, and how Congress and the American public need them.

Increase Accountability: The Library will ensure AI implementations are transparent and are used to augment staff responsibilities to serve Congress and the American people.

Foster Equity: The Library will proactively identify and reduce potential harms from biases and imbalances inherent in data-driven AI technologies by collaborating with communities and setting high quality standards for equitable outcomes.

Ensure IT Security: Any implementation of AI will appropriately and thoughtfully consider information technology, IT security and privacy risks, and take steps to mitigate those risks.

Protect Rights: AI implementations will protect the intellectual property, privacy, civil liberties and civil rights of Congress, staff and the American people.

This definition and set of principles represent thinking across the Library about the data, practical and ethical aspects of these technologies and will guide our work as we continue to navigate the world of AI.

Approach and Innovation

Strategic thinking and digital transformation pave the way for the adoption of AI technologies. The Library’s FY2024-28 Strategic Plan, A Library for All, strategically integrates digital strategy and approaches throughout, embedding digital capabilities into all of our goals and objectives. The plan includes a bold vision for exploring and incorporating new technologies and methods to better serve Congress and the American people. As the plan states, the Library is building upon its current approaches, collaborations and mechanisms to continuously assess the benefits and risks of emerging technologies, plan for their adoption and expand governance to ensure reliability, efficacy and ethical use.

The Library has a long history of exploring new technologies and cultivating standards in information sciences. Our responsibility to be a trustworthy steward of information requires both caution and experimentation. At the same time, our responsibility to efficiency in serving our mission, and our position as the largest library in the world, means the Library is a source for standards and best practices in adopting new technologies and methodologies across multiple sectors.

The Library explores and experiments before implementing AI technologies. The Library’s ML and AI experiments are critical to our strategic approach in embracing and adopting these technologies. These explorations are collaborative, test AI in our unique collections, integrate other technological advancements we have created or adopted, and assess the risks and benefits of these technologies for the Library and our user communities. The Library’s Digital Innovation Division (LC Labs) has conducted AI and machine learning experiments addressing the ethical and practical dimensions of using AI tools in libraries and cultural memory organizations. These experiments include testing off-the-shelf tools, publishing state-of-the-field reports, and conducting research explorations into how ML and AI can be applied to the Library’s collections, workflows and public services. A planning framework created by LC Labs supports the responsible exploration and potential adoption of AI at the Library, including three planning phases and worksheets, questions, and workshops.

Our implementation of AI is rigorous and responsible. As complex uses of AI and ML technologies surface, the Library is building on our continuous work with emerging technologies. We review and implement AI use cases in alignment with our values of trust and authenticity, taking into consideration bias, predictability, accuracy, IT security, intellectual property rights, and ethics.

How We Are Using Artificial Intelligence

The Library has been using forms of AI and ML technologies for years. For example, optical character recognition (OCR) technology used to process documents and enable full-text searching has taken advantage of machine learning methods for over 10 years. This technology is employed across our text collections, including the Chronicling America collection of historic American newspapers.

The Library’s current uses of AI includes experiments, pilots, tools and implementations across the Library. Our AI use cases employ a focused, “humans in the loop” approach consistent with the Library’s core work providing research and information to Congress, stewarding our nation’s vast trove of library and archival materials and administering the United States Copyright system. We are currently using AI to:

  • Extract data from historic copyright registration records;
  • Geotag legislative record content;
  • Create machine-readable maps;
  • Segment image and audio files;
  • Enhance reader recommendation and written correspondence services;
  • Enhance legislative data workflows; and
  • Monitor performance and increase IT security services.

To find out more, including use case names, summaries, and evolution of AI use cases, please consult the full Library of Congress AI Use Cases page [Direct link coming soon].

About the Digital Strategy Directorate

The Digital Strategy Directorate leads the Library’s strategic digital efforts with a focus on emerging methods and technologies, such as AI, that better serve the Library’s mission and our communities. It also manages the Digital Innovation Division (LC Labs) program.