This 45 minute webinar runs through the process I use to "attack assessments". It begins with a review of materials to check whether the assessment is first valid, and then whether it is vulnerable to GenAI misuse.
I discuss assessment validity through the lens of three Australian curriculum authorities, and outline some basic checks for our assessments before we worry about AI: Is the assessment evaluating what we mean to assess? Are we using the best available assessment methods? Have we made sure the assessment happens over time, and in a variety of informal and formal contexts?
Finally, we incorporate the AI Assessment Scale and the "Swiss Cheese" model to decide when and how AI might suit the various tasks which now form our valid assessment process.