The National Library of Australia has released an AI Framework to help it explore the benefits of AI while minimising risks.
The Artificial Intelligence Framework released by the National Library of Australia (NLA) in February 2025 will see "careful deliberation and measured enthusiasm" (p 3) backend AI adoption at the Library. While the document is repetitive in parts, it does set out useful thinking for GLAMs and other organisations when consider their own responses to AI.
The NLA will use AI to extend "the Library's practices to support its statutory role of maintaining, developing and protecting a national collection of library materials so Australians can discover, learn and create new knowledge, now and in the future" (p 3). Leveraging emerging technologies like AI are designed "to engage Australians with their history in ways that present historical and contemporary information in more interactive and enlightening formats, making the past more relatable and accessible" (p 13) and "to foster deeper connections with Australia's heritage, encouraging reflections and new conversations" (p 13).
Don't just read the beginning. Become a paid subscriber and you can read the entire briefing. Subscribe to keep reading.
The NLA will anchor AI to legal, responsible and ethical uses, backed by guardrails and governance, that support the Library's mission and statutory function (p 3). But, in adopting AI, as with any new technology, the NLA's legal and statutory obligations will not be compromised (p 3), nor will the Library's trusted and respected reputation (p 3). This includes being responsive to evolving regulation in areas such as privacy, copyright, Indigenous Cultural and Intellectual Property (ICIP), Indigenous Data Sovereignty (IDSov), inclusion, accessibility, digital collecting and cyber security risks (p 3).
The vision that guides the NLA's use of AI is to "legally and responsibly innovate with AI to ethically unlock the national collection, amplifying its discoverability, breadth, accessibility and impact in ways that do not perpetrate or entrench harms" (p 3).
