Jena Zangs, chief data and AI officer at the University of St. Thomas, spoke with University Business about different methods to mine AI for insights safely. As AI emerges and becomes a tool used in academia, Zangs noted that AI users should summarize documents and records containing sensitive information before uploading the information for analysis by AI platforms and chatbots.
From the article:
Different versions of ChatGPT, Microsoft Copilot and other popular chatbots process user data in vastly different ways.
Institutions lose control over data if it’s uploaded into free, public models of an AI chatbot. Enterprise models, on the other hand, have custom settings that dictate how information is stored and processed and which users have access. ...
By obscuring personal information, this strategy prevents administrators from, for example, uploading “a student’s grades that someone else could benefit from,” says Jena Zangs, chief data and AI officer at the University of St. Thomas in Minnesota. ...
“AI is used to strengthen our human capabilities,” Zangs says. The culture around adopting AI at St. Thomas is “grounded in the human voice.”
