Turnitin releases its first Learning Integrity Insights report on the use of artificial intelligence in education, finding that educational institutions are moving away from AI detection alone and toward classroom integration and customization.
Based on global qualitative and quantitative data from educators, students, and institutions, the report points to changes in the way schools and universities manage AI. Many educators are now looking for clearer information about how and why AI was used in their work, rather than just focusing on whether students used it.
This reflects a broader shift away from a simple “yes” or “no” approach to AI in the classroom. More than 60% of recent customer feedback prioritizes transparency in AI usage, while educational institutions that have already begun integrating AI are increasingly asking for ways to tailor approved usage to specific classes and assignments.
Annie Chechitelli, Turnitin’s chief product officer, said the company recognizes the demand for a more flexible approach.
“There is no one-size-fits-all approach to what responsible use of AI in education looks like,” said Annie Chechitelli, chief product officer at Turnitin.
This comes as many institutions are still developing formal rules regarding AI. From our discussions with customers, less than half said they had an AI policy, despite evidence that clear and consistent rules benefit both educators and students.
policy gap
The lack of a formal policy appears to be a major barrier to broader and consistent adoption. With rules in place, staff and students will have a clearer basis for deciding when the use of AI is permissible and how it should be disclosed. When these are missing, uncertainty remains about the role of AI in education, feedback, and assessment.
The report also highlights feedback gaps. Teachers say they often don’t have time to give students as much feedback as they want, and students may not have enough opportunities to seek guidance during the writing process. AI could help close that gap, but hesitancy remains about using it for feedback and grading.
Checchitelli said hesitancy largely depends on the context.
“Working with educational institutions on solutions that support the goal of responsible AI use is yielding insights into what works and what doesn’t. What we’re hearing from educators is that those who are incorporating AI into education want more customization and insight into how students can use AI, and AI in feedback and grading. We’re also hearing from both educators and students that their comfort with the presence of a child is situational,” Chechitelli said.
Custom applications
The emphasis on customization represents a notable step in the conversation surrounding AI in education. Early concerns centered around detection and fraud, especially after generative AI tools became widely available. This new research suggests that more and more educational institutions are accepting the presence of AI in the classroom and are now focusing on guardrails, disclosures, and assignment-level rules.
That doesn’t mean old forms of academic misconduct have disappeared. Despite the shift in attention to AI-generated content, traditional plagiarism still persists. Since April 2023, 6.16% of posts have 80% or more similarity to existing sources. Of the highly similar papers, 91 percent matched primarily to other students’ papers, 7.8 percent matched to Internet sources, and approximately 1 percent matched to publications.
This pattern is very similar to the period before generative AI tools were more widely released. From the end of 2019 to the end of 2022, 6.86 percent of posts had 80 percent or more similarity. Of these, 88.25 percent matched primarily with other students’ papers, 10.75 percent with Internet sources, and approximately 1 percent with publications.
These numbers suggest that while AI has changed the conversation around academic integrity, established plagiarism patterns remain relatively stable. For educational institutions, this may reinforce the need to manage both issues at once, rather than treating AI as a replacement for previous concerns about copied works.
This report is based on customer discussions about how AI fits into learning, assessment, and scoring. Institutions are increasingly looking for practical ways to balance experimentation and contrast, especially where faculty need the space to set different expectations for different subjects and assignments.
Turnitin works with more than 16,000 customers in 185 countries and territories. Less than half of institutions report having an AI policy.
