Guidance and Governance for AI’s use in Research (AI4Res)

The emergence of generative AI is perhaps the biggest challenge facing research development planners at HEIs and other organisations with substantial research facilities and needs. It has also presented a serious policy and governance challenge in regards to research ethics, data protection and management and general research practice.

It is critically important that research development training in the AI4Res and university policy governing the use of AI align. This dovetailing must be led first and foremost through a commitment to research ethics and integrity and, through this, compliance with legal requirements and research standards.

I have developed both RDT and high level HEI policy together at a UK HEI. This relied on a number of sources, but particularly UKRIO guidance – namely its Embracing AI with in Integrity paper – and the policy and guidance of a HEI that acted as accreditor for my university’s research degrees.

Here, some of the best examples of HEI AI guidance and policies are highlighted and discussed in regards to developing good practice researcher training in the area of AI. Special read is also made to the Vitae Research Development Framework.

Vitae and UKRIO

Firstly, the rapid and enormously impactful emergence of AI raises some serious questions about all approaches to researcher development and researcher ethical practice capabilities especially. Point 2.4. of the Vitae’s ‘knowledge and practice’ descriptor of Vitae’s RDF highlights the ‘digital and technological capabilities’ or researchers. No longer limited merely to ICT skills and capabilities regarding software use, data management and storage and simple MS Office tools, researchers must now learn and enormous amount about how to use AI for the following reasons:

  • How to use it productively and ethically themselves in their own research
  • How to approach its use by colleagues, with an eye to safeguarding good practice and ethical research culture as well as perceived competitive dis/advantage with peers
  • How to approach its use as ethical reviewers or peer reviewers
  • How to approach AI use with regard to the policies of academic journals and funders

From a research ethics standpoint, there are also serious implications for the Vitae RDF in regards to points 2.5, 2.6 on critical thinking and creativity in the same ‘knowledge and practice’ descriptor as well as 3.5 on research ethics and 5.1 on research culture. In truth, the implications for AI’s emergence for every part of the research cycle mean that that most of all existing research development framework’s and training programmes need to be radically rethought.

UKRIO offers research training to subscriber organisations (mostly, but not exclusively universities). Its Embracing AI with in Integrity report provides a balanced set of facilitatory guidance, information as well as clear red lines for ethical practice.

In devising training programmes on AI for researchers, it is important this is tied to the fast changing governance and information systems that univ

References and Sources

Leave a Reply