When carrying out a work task, you must always ensure that you have sufficient professional expertise to assess the work and ensure that it is completed thoroughly and to a standard acceptable to all stakeholders.

You must be aware that certain types of work are subject to specific requirements, including those relating to documentation, transparency and impartiality. This means that the use of AI must meet additional criteria before it can be applied to such work. This is particularly important in situations governed by specific regulations designed to protect individuals’ rights in decision-making processes.

If you become aware that you or a colleague has used a platform in breach of the rules governing the performance of work tasks, or that a security breach has occurred, you must always report any inappropriate, incorrect or unauthorised use to your immediate manager and follow AAU’s reporting procedures.

For personal data that is not lawfully available to the public, you may only share it with platforms that have been approved for processing that type of personal data.

Data that does not relate to individuals may still require special protection. This includes, for example, data entrusted to us and information about new, groundbreaking research findings. You must only share such data and information with platforms that have been assessed as providing adequate protection. The University’s overview of data classification specifies which systems have been approved.

Be particularly mindful that the use of generative AI may raise copyright issues. This applies both when copyright-protected material, such as text or images, is shared with a platform and when generated output may infringe copyright.

If you generate text, images, video or other content, you should always ensure that its use of AI is documented. For example, add ‘AI-generated using ChatGPT’ to image filenames and save the prompts and responses together with the output, such as in a separate text file, in PowerPoint comments or by similar means. The important thing is to ensure that anyone who later accesses the material can see that it was generated using AI.

You must always clearly label content as AI-generated if recipients could now or in the future mistakenly believe it to be authentic, such as photorealistic AI-generated images. This is essential to maintaining trust in AAU.

Some areas have specific rules, standards and practices for declaring and documenting the use of generative AI. This typically applies when creating content to be shared with third parties, such as funding bodies or news organisations. Always check whether they have specific requirements for how you may use generative AI to create content.

You understand the strengths and limitations of the platforms you use and apply critical thinking when evaluating their results.

When carrying out your work, remain mindful of AAU’s commitment to promoting inclusion, equality and diversity and preventing discrimination.

Use your judgement, recognising that these platforms often lack contextual understanding and may fail to account for nuances in particular circumstances.

When carrying out our work, we always ensure that anyone with relevant expertise, specific responsibilities or a particular role is appropriately involved, as are those who may otherwise be affected by a new approach. We foster a psychologically safe environment in which we can discuss professional aspects of our work and respect the expertise that each of us brings.

You respect the time of colleagues and others by continually assessing the relevance and value of using AI-generated content.

We continually consider the climate impact of our use of these platforms and assess how our choices can reduce their negative environmental effects. We also consider how the platforms can support more climate-friendly approaches to administrative work.