PULLMAN, WA – As generative artificial intelligence is woven rapidly into society, teachers-in-training — as well as the professors who educate them — feel unprepared to adopt the technology in the classroom, according to a survey conducted by a Washington State University instructional designer.

To respond, institutions should implement clear guidelines and provide professional development opportunities for educators, the author of the new paper says.

While faculty serve as subject matters experts who best understand their course needs, they require institutional support and learning opportunities to grasp how these algorithms work, their affordances and limitations, appropriate usage, and ethical considerations. This foundational knowledge will help educators make a thoughtful decision about integrating these tools or not into their courses.

“The main takeaway of all of this is that our students are asking to learn more about AI, our teachers are asking to learn more about AI, and we do not have the support to do it,” said Priya Panday-Shukla, an instructional designer in the WSU Global Campus whose paper will be published in the October issue of Teaching and Teaching Education but is available online now.

“The whole point is that if faculty have that information, they will be able to make an informed decision: ‘OK, this is what I teach, this is what I do, and maybe I could use it for this or for that — or  maybe I should not use it,’” she said.

Panday-Shukla’s paper surveyed two populations in early 2024: Fifty-two pre-service teachers with an average age of 20, most of whom were set to graduate in 2025, and 21 teacher educators with an average age of 54, most of whom had taught at the university level for many years. She also conducted follow-up interviews with some of the participants.

Though Panday-Shukla’s survey showed a range of attitudes toward GenAI, majorities of both pre-service teachers and university educators said they had not received any training for how to implement it in their own classrooms.

Forty-eight of the 52 students said they hadn’t used AI as part of their current classes and 49 said they had received no training on how to use it in their own teaching. Among professors, 18 of 21 said they weren’t using AI in their classrooms and had not received training on how to use it.

Panday-Shukla said that an unfamiliarity with AI might lead some professors to avoid using it altogether. But that does a disservice to students who will graduate into a world being rapidly transformed by the technology. One estimate shows that some 30% of hours now worked nationwide could be automated in five years, and millions of workers will likely need to seek new fields.

The whole point is that if faculty have that information, they will be able to make an informed decision: ‘OK, this is what I teach, this is what I do, and maybe I could use it for this or for that — or  maybe I should not use it.

Priya Panday-Shukla, instructional designer
Washington State University Global Campus

Panday-Shukla has developed a workshop for educators through the Global Campus that provides a framework for considering how to use GenAI. She’s hoping to expand the workshop across campus. Using the OSPI’s AI matrix for K–12 classrooms as a starting point, she developed a system that would establish four graduated levels of AI use in a classroom — ranging from a flat prohibition on its use to a requirement that students use GenAI. Educators can use this framework to be transparent and provide clear guidance to students about the levels of assistance allowed for different assignments and the necessary steps students should take when utilizing these technologies.

Panday-Shukla emphasizes that the technology is not a replacement for original research or writing, or a shortcut for academic rigor. “When you need to verify information, you still do it the old way,” she said. “You check it one source at a time, one piece of information at a time. It’s no different from that.”

Her own paper demonstrates one of the uses of GenAI — as well as how to be transparent about using it. As she acknowledges in a declaration at the end of her publication, she used Google Gemini to check some passages in the paper for clarity and readability. She used the resulting proposals as suggestions, not cut-and-paste solutions. Such declarations are becoming common requirements in journals.

“It’s just another tool that requires thoughtful integration, and therefore, we have to learn how to use it properly,” she said.