Here are suggestions for how instructors may adapt their assessments to maintain validity even as generative AI grows in prevalence and sophistication.
Shift emphasis to in-class assessments
Consider converting some take-home assignments to in-class assessments without internet access. Larger assignments may be broken up into component pieces. For example, as part of a larger group project, require students to write a 500-word draft of a report section as an in-class assessment.
Include AI-resistant formats in the mix of assessment formats used
If appropriate, complement and vary different forms of assessment. For example, follow up a written assignment with an oral presentation.
Require personal insights or contextual information
Use questions that require information that AI does not have access to, such as personal experience, or include the requirement for a reflective response, which builds in personal insight. AI tools are far less useful in this context.
Introduce Authentic Assessment
Involve tasks or activities that closely mirror real-life challenges, requiring students to apply knowledge in a practical, context-specific problem-solving context. For example, in certain law courses, students can be assessed using mock trials.