AI Detectors: At best, buying us a little time to get our act together? – Dr Robin Crockett
15.00-16.00 6 Mar 2025
Part of the Academic Integrity online webinar series Feb/Mar 2025
AI ‘writers’ – chatbots – continue to improve and the text-quality continues to get closer to human equivalence, posing even more questions about AI-text detectors. When human equivalence is attained, AI detectors will be faced with the impossibility of distinguishing between typical/aggregate human text and AI-generated text which precisely matches it.
This presentation will focus on this progression and what we can do in the meantime:
- Continue to refocus student assessment away from things AIs do well towards things where human insight and understanding are required…while that remains possible…
- Continue to observe reliable indicators of AI misuse
- Get our act together for the AI era
