This blog post is inspired by an event chaired by Simon Buckingham Shum with presenter Paul Stacey during UTS’s Open Education Week. For more detail, view a full recording of the event (65 mins).

At first glance, the possibilities offered by Generative AI tools and Open Education seem broadly aligned. Both are making content available to learners and educators who might otherwise not have access, reducing costs and opening up new possibilities for educational content creation and sharing. Look more closely, however, and the tensions between them begin to surface.

GenAI and open education in conflict?

Most of the licences that are available to make something open – whether it’s source code or educational resources – rely on intellectual property law or copyright law, and AI is really treading on tricky ground there…

Paul Stacey

The 5 Rs of Open Educational Resources (OERs) are made possible by a series of Creative Commons open licences, balancing the assertion of copyright with permission for others to build on creative content. This process supports a range of benefits associated with OERs, from affordability and access (for students and institutions) to diversity in educational materials, cross-institutional collaboration and improved learning outcomes.

So what happens when content is generated in completely new ways by GenAI, where authorship is unclear, the source of data and content cannot be traced and sharing permissions are up for debate? As Paul Stacey noted during Open Education Week, there are many implications for how open resources have traditionally worked within education – and these are potentially in conflict with what’s happening across GenAI.

The AI technology stack: implications for open and education

Paul framed the impact of GenAI on open education by talking about the entire AI technology stack, from the hardware and cloud platform layers at the bottom, through the data inputs, ingestion and training, models, APIs, AI apps and applications, to the user interactions and generative AI outputs. The way things are done through each of these layers raises questions or challenges for open education, such as:

  • Who owns the outputs of GenAI? Does copyright or commons licensing apply, are outputs deemed to be in the Public Domain or owned by the provider of the AI tool in question?
  • Who owns GenAI prompts and questions? Can these be licensed?
  • Should AI apps and interfaces be built on open source software?
  • How do proprietary vs open models work? What are the implications for content generated through a closed or proprietary training model?
  • Is the use of ‘scraped’ data always ethical or even legal? If data sources are not disclosed, what does that mean for quality, transparency and trust in open content?

These questions and more continue to be debated at the same time as GenAI is being used and designed into educational processes from learning design to assessment. With pressure from government, industry and the workplace to keep up with these technological advances, educational institutions and the resources they use and generate (including OERs) must remain relevant.

Where to from here?

The world of AI is at an important crossroads. There are 2 paths forward: one where highly regulated proprietary code, models, and datasets are going to prevail, or one where Open Source dominates. One path will lead to a stronghold of AI by a few large corporations where end-users will have limited privacy and control, while the other will democratize AI, allowing anyone to study, adapt, contribute back, innovate, as well as build businesses on top of these foundations with full control and respect for privacy.

The AI renaissance and why Open Source matters – Nick Vidal, Open Source Initiative

With GenAI’s potential to generate resources and content, from learning objectives to course materials, assignments and feedback, where does this leave OERs and open education in general?

Paul Stacey’s call to action highlighted a number of areas we might choose to focus on in open education, including:

  • Learning, mapping and experimenting – creating and testing new frameworks as the tools evolve
  • Creating guidance and guardrails for students, educators and institutions
  • Being part of the evolving conversation through policy and advocacy
  • Collaboration across institutions

It seems to me that open education is ultimately about access, sharing, recognition and building on the work of others, with OERs enabling learning and pushing the boundaries of knowledge in ways that include everyone. If GenAI tools do not support these goals, we might ask ourselves what is at stake, what might be lost, and what it means for the broader goals of education for all.

Learn more about GenAI from an open education perspective in this long-form piece by Paul featured on his blog.

Join the discussion