Artificial intelligence (AI) has already moved into classrooms across the country, but the people responsible for guiding its use say they haven’t been given the tools to do so. That’s the central finding of a new survey from Instructure, the education technology company behind the Canvas learning platform, released this week during the company’s annual InstructureCon conference.
The survey, conducted in mid-June 2026 through SurveyMonkey Audience, gathered responses from 1,125 people across three groups: 410 K-12 and higher education instructors, 312 recent or current college students, and 403 parents of K-12 children. It represents the opening phase of what Instructure describes as a longer-term research effort to track how AI is reshaping schools and universities.
The headline gap is stark. Ninety percent of college students say they use AI in their coursework at least occasionally, and nearly three-quarters of parents report the same about their children. Instructors are using the technology too, though less often — 68% of K-12 teachers and 61% of higher-education faculty report at least occasional use.
Yet formal preparation hasn’t kept up. Roughly four in ten educators in both K-12 and higher education said they’ve received no formal training on AI at all, while only about one in ten described their training as comprehensive. That gap shows up in how educators rate their own confidence: on a five-point scale, K-12 teachers averaged 2.53 and higher-education instructors 2.60 — both below the midpoint.
Despite the shortfall in training, outright hostility toward AI appears rare. Large majorities across all three groups — including 94% of college students and 88% of parents — cited at least one reason for optimism about the technology’s role in education. Parents pointed most often to accessibility gains for students with different learning needs, while educators tended to value the time AI saves them.
That optimism, however, comes bundled with wariness. A majority in every group — from 60% of K-12 teachers to 69%of parents — worried that AI tools can present wrong information with unearned confidence. Respondents across the board also flagged concerns about students leaning on AI too heavily, losing critical-thinking skills, or using it in ways that blur into plagiarism.
That tension appears to be resolving into a rough consensus about where AI belongs. Survey respondents were considerably more comfortable with AI being used to explain concepts, brainstorm ideas, or help locate resources than they were with AI playing any role in grading or academic decision-making. Teaching students to use AI responsibly ranked as a higher priority than restricting their access to it outright.
“AI use is already part of how students learn and educators work,” said Melissa Loble, Instructure’s chief learning officer, adding that the priority now is ensuring people have adequate training and clear guidelines. Mary Styers, the company’s director of evidence and learning strategy, said the notable finding was the degree of agreement across otherwise different groups: all three want clearer limits and better support, not a retreat from the technology altogether.
Instructure said further findings from the research initiative will be released in the coming months.
