Annie Chechitelli is the Chief Product Officer at Turnitin, where she oversees the Turnitin suite of applications, including academic integrity, grading and feedback, and assessment capabilities.
Turnitin is a global company dedicated to ensuring the integrity of education and meaningfully improving learning outcomes. For 25 years, Turnitin has partnered with educational institutions to promote honesty, consistency, and fairness across all subject areas and assessment types. Turnitin products are used by educational institutions and certification and licensing programs to uphold integrity and increase learning performance, and by students and professionals to do their best, original work.
How has the introduction of AI writing tools like ChatGPT impacted academic integrity?
ChatGPT’s worldwide accessibility has created new potential outlets for academic misconduct, including AI generated writing, plagiarism, and paraphrasing, at a massive scale that only continues to grow. For educators, the ability to distinguish between AI and human generated content is only getting more difficult. As a result, ensuring students are creating their best, original work goes far beyond detection. Educators need the ability to intervene before misconduct occurs in order to reinforce academic integrity and ethical writing practices. AI isn’t going away, and the new age of academic integrity requires a robust understanding from both students and teachers on how AI can be used responsibly to support human learning and decisions.
At Turnitin, we’re committed to helping students and educators alike understand the role of AI in education, including identifying misconduct issues and assessing how it can be best integrated into the classroom. In viewing AI writing in education through the lens of just one puzzle piece – identifying AI writing in students’ submissions – institutions and educators are missing the full picture, because AI can benefit learning as well. This technology has the ability to augment human ideas and facilitate a greater understanding of complex topics if used correctly. AI itself isn’t a threat to original thinking or academic integrity as long as students know how to use the tools ethically. Its educators responsibility to help provide their students with the guidance on what is and what is not a permissible use of the technology – a skillset that will help them beyond the classroom and into their future professions.
What specific challenges do educators face in detecting AI-generated and paraphrased content?
Higher education students’ use of AI far outpaces educators’ use of the technology. Turnitin partnered with Tyton Partners* and found that of the students surveyed, 59% say they use AI writing tools at least once a month, while as for administrators 40%, and 36% for instructors. A lack of familiarity with these tools leaves educators in the dark on AI’s abilities and limitations.
Additionally, AI use is not just about AI-generated content anymore. AI paraphrasing has taken off and is the latest method some use to avoid detection. Educators need support tools that keep up with AI’s latest advancements and provide insight into a student’s workflow and thought process.
These features are not meant to replace the educator’s professional discretion, but rather to supply additional insights to help reach a decision. Reports, such as ours, provide data points and resources to support a process of working with students, not determinations of misconduct.
Instead, our AI writing and paraphrasing tools are designed to support educators in initiating constructive conversations with students about the importance of ethical writing practices and the proper use of paraphrasing techniques.
We’re also building a new solution for student writing to give both educators and students confidence in the authenticity of work by focusing on “proof of process.” Coupled with our existing plagiarism and AI checks, our goal with this solution is to give educators confidence and insight without burden, and offer students comfort in knowing their hard work won’t be judged solely by a score. That product will be used by a select few universities later this year.
*Turnitin was a partner in providing compensation to conduct Tyton’s Time for Class 2024: Unlocking Access to Effective Digital Teaching & Learning.
Can you explain how Turnitin’s new AI paraphrasing detection feature works?
AI paraphrasing refers to the process of using AI technology to rewrite text while retaining the original meaning. These AI-powered tools, also known as text spinners, analyze the input text and generate alternative versions that convey the same information using different words or sentence structures. The goal – to make AI-generated text go undetected.
However, Turnitin’s AI paraphrasing detection feature identifies instances where AI paraphrasing may have been used to modify AI-generated text in submitted content, helping publishers and educators to maintain the integrity of research writing and foster authentic learning experiences for students.
Within the platform, the AI writing report highlights the text segments that the model predicts were likely written by an AI tool. Separately, it also highlights AI-generated text that may have been further modified using an AI paraphrasing tool, so educators and publishers can understand the nuances of detection.
How have your previous roles at AWS and Amazon, particularly in education technology and digital content distribution, influenced your approach to developing and enhancing Turnitin’s AI paraphrasing detection feature?
Throughout my career, I have always focused on the customer-first mentality: constantly evolving, responding to the market and identifying customer needs and feedback. This is a mentality I have carried with me into my role at Turnitin and is at the heart of our product development strategy.
Technology has always changed at a rapid pace and the same can be said about AI. Following the launch of our AI writing detector, Turnitin heard from educators around the globe that students were trying to bypass detection by using paraphrasing tools. Students and researchers may have used AI paraphrasing to modify AI-generated content, including content produced by language models like ChatGPT, in an attempt to evade detection by AI detection software.
In turn, we developed our AI paraphrasing tool in July of 2023, seeking to address this issue by identifying instances where AI paraphrasing may have been used to modify AI-generated text in submitted content. Our goal is to help publishers and educators maintain the integrity of research writing and foster authentic learning experiences for students, especially ahead of the 2024-2025 school year.
In what ways can Turnitin Originality help institutions maintain high standards of academic integrity amidst the rise of AI writing and contract cheating?
Turnitin Originality has set an entirely new standard for academic integrity for both students and educators. It goes beyond plagiarism to address new and changing forms of misconduct, like AI writing, contract cheating and paraphrasing, and provides educators with clear, actionable data for tracking student performance across their cohort and assessing potential misconduct issues.
Students also receive support through real-time formative feedback on their work. Originality compares text similarity, grammar, and citations as students work, helping them make corrections and avoid unintentional plagiarism before final submission.
What feedback have you received from educators and institutions about the effectiveness of Turnitin’s AI writing and paraphrasing detection tools?
Recently, Turnitin sat down with Dr. Leslie Layne, an associate professor at the University of Lynchburg, Virginia (USA). Layne has taught 76 freshman writing classes and has used Turntitin since starting at the university in Fall 2004.
Reflecting on the challenges before the advent of efficient AI writing detection tools, Layne recalls, “It was really time-consuming to copy and paste text into online detectors.” This process, while necessary, was far from efficient and 100% accurate, underscoring the need for more streamlined and reliable solutions. The introduction of Turnitin’s AI writing detection tool was a game-changer, providing consistency and a much-needed ease of process.
“Grading essays takes time. I’m very fine tuned to the amount of time that it takes to get done what I need to get done, so I was absolutely delighted to hear that Turnitin was incorporating AI writing detection, because then everything is in one place,” said Layne.
A central theme in Layne’s approach is the constructive use of AI in education. Rather than viewing AI as a threat to academic integrity, she advocates for its integration into the learning process. “I do teach with AI in my freshman writing classroom, so I’m having them use ChatGPT in almost every class period,” she explains. This hands-on approach with AI equips students with the skills to navigate and leverage these tools both effectively and ethically.
How does the Draft Coach™ feature integrate with Turnitin’s other tools to support students in producing original work and improving their writing skills?
The Draft Coach™ feature exists within the Originality product. Just like Originality supports teachers in upholding academic integrity, Draft Coach provides students with guidance on how to improve their academic writing and research skills at all times. Instant feedback on text similarity, grammar, and paraphrasing helps them avoid accidental plagiarism.
What role do you see AI playing in the future of academic integrity, and how is Turnitin preparing to address potential new challenges?
AI writing tools have the potential to make our lives easier, including enhancing and tailoring content, bridging language barriers, increasing productivity, and assisting with brainstorming. However, before someone can reap the benefits of these tools, they need to understand how to use them responsibly and ethically. The classroom can be part of the solution. It is a safe space for students to learn, practice, take chances, and grow with technology. The lecture hall can serve as a great training ground for students to understand its ethical boundaries, safely experiment, learn to discern between facts and inaccurate information, and strengthen critical thinking skills.
To help educators and institutions navigate the new reality, they need trusted partners, like Turnitin, who understand their needs and those of their students. In fact, two years prior to the public launch of ChatGPT, Turnitin began working on how to provide educators with a tool that gives them transparency into the use of AI writing so they can guide students to use AI responsibly and support academic integrity.
As AI evolves, Turnitin’s products will too. New features can be easily integrated into existing products and workflows for ease of customers’ use and a decreased learning curve. We regularly provide updated student, educator and institutional resources to help them navigate this new technology and sentence-level detection data.
Finally, a number of our employees have a background in education, which ensures our products are pedagogically sound. Without their insights, we wouldn’t be able to meet the evolving needs of educators and administrators. With all of this in mind, our upcoming “proof of process” solution, which will be integrated into existing Turnitin workflows, will give both educators and students confidence in the authenticity of work.
What advice would you give to educators and institutions looking to enhance their academic integrity policies in the face of evolving AI technologies?
My opinion on academic policies is that having multiple AI policies at different institutional levels is a key way to ensure situational nuances are accounted for. For example, an AI policy for a creative writing class may not be the best way to approach potential AI use on a research paper. Instructors creating assignment-level policies should provide details on acceptable AI use based on the assignment’s learning objectives.
At the institutional level, however, academic integrity policies should be broad with a focus on how students represent their work and cite their sources. This prevents intentional plagiarism while giving space to nuances associated with the unique situation.
Across the board, it’s crucial to look beyond a single detection percentage. Our tools are designed to support institutions and educators, not replace them entirely. Teachers know their students best, and should use any detection of AI to facilitate productive conversations about appropriate AI usage and critical thinking.
In order to best support educators, Turnitin also has a vast library of resources and an expansive teacher network available to help institutions build out their policies, including a readily available guide that shares detailed insights on how to review and revise existing academic integrity policies to address threats from AI, available online to download free of charge.
Thank you for the great interview, readers who wish to learn more should visit Turnitin.
