Recommendations from the joint VU-UvA taskforce AI in education

Recommendations

• Technical, legal and ethical AI checklist VU and UvA

  There is a wide variety of providers of AI applications, ranging from the well-known big tech companies to open source communities. There has long been a question within the VU and UvA as to how they should relate to these providers of digital technologies. The digital sovereignty of the institutions is a relevant discussion point when it comes to choosing an AI vendor.
  
  When choosing a provider of AI applications, the VU-UvA task force suggests using a checklist that includes considerations that secure ethical, legal and technical use of these applications. This is not only a compliance checklist, but also a document that encourages critical evaluation.
  
  The purpose of this checklist is to communicate the strategic positioning of the VU and UvA towards AI providers. However, it is important to realise that this checklist is not entirely ‘finished’ and will never be finished, because developments do not stand still. It remains relevant to further sharpen the checklist in case of major developments and align the requirements accordingly.

  Read the comprehensive technical, legal and ethical AI checklist here.

• Preconditions for exploratory research on AI in education

  So far, little research has been done on the opportunities and risks of AI in the education (process). However, we do know that AI is already widely used by students and staff. This makes action desirable to both anticipate the risks and capitalise on the opportunities that AI can begin to provide for education.

  Research and teaching are the core tasks of academic institutions. The task force advises both institutions to encourage (applied) research in the field of AI in education. In addition, the task force advises to actively engage in evaluation of exploratory research to identify best practices with which the institutions can further shape AI in education. The checklist (see above) provides frameworks for experimentation.

  Read the comprehensive recommendations on the preconditions for exploratory research on AI in Education here (only available in Dutch).

• AI literacy

  Due to the rapid advance of generative AI, and thus lack of research into long-term effects, it is still unclear how generative AI can be deployed in a sustainable way

  can be used in education (looking at educational quality). The UvA and VU should therefore guard against thoughtless deployment of generative AI in education. In any case, knowledge and skills are needed to work optimally with (generative) AI tools: AI literacy. This refers to artificial intelligence in general: not just generative AI.

  In the attached document, some learning objectives and starting points are elaborated, to concretely translate AI literacy into educational practice. The task force recommends starting by encouraging teachers to professionalise in the field of AI literacy, so that they can make critical judgements about where to allow and where not to allow AI in their teaching. If they allow it in a certain way, it is important that they can support their students in using AI in line with ethical values and their discipline. The task force recommends that every teacher and student should be ‘AI literate’ regardless of whether they themselves use AI (in teaching)

  Read the comprehensive recommendations on AI literacy here (only available in Dutch).