Turnitin AI

We use state-of-the-art artificial intelligence to help students, educators, and institutions advance integrity and learning outcomes, at scale.

At Turnitin, we believe in human-centered AI. This means putting the human user at the center of research, development, and function. We seek to build AI that assists, strengthens, and scales human abilities—not replace them. Our work with AI furthers our mission to ensure that students, educators, and institutions have the power to make data-driven decisions that promote academic integrity and improve learning outcomes.

Eric Wang

VP, Artificial Intelligence

Turnitin AI

Our guiding principles for AI

AI should be built with rigorous and peer-reviewed statistical and machine learning standards.
AI should be designed, developed, and tested by a wide and diverse range of learners and educators, not only engineers and AI scientists.
AI should be continuously improved to make it more accessible, fair, and beneficial to learners and educators.

AI is only as good as the principles on which it is built.Turnitin is steadfast in its commitment to responsible and transparent AI.

AI should help improve learning outcomes and promote and protect academic integrity.
AI should be designed to adhere to high standards of privacy and data ownership.
AI should be intentionally designed to mitigate the impact of unintended biases.

How we use AI

Give instructors tools that streamline manual grading processes, so they can spend more time teaching.

Answer-grouping

AI-assistance helps make grading easier, more efficient, and more fair by allowing instructors to score and give feedback on groups of similar responses at once.

Handwriting recognition

Some assignments and exams include handwritten calculations, codes, diagrams, proofs, and more. Turnitin AI technology identifies and transcribes these responses to take the administrative work out of grading.

Authorship verification

Advanced forms of plagiarism are difficult to identify from contract cheating to word spinning and bot writing. Using AI to compare writing styles, Turnitin insights help indicate authorship. Authorship profiles use forensic linguistics and bot writing identification uses detectors built from models such as GPT-3.

Document analysis

With AI, written and hardcopies of exam templates can be converted into electronic tests, helping streamline assessment. We combine vision-based machine learning with PDF parsing techniques to perform document understanding tasks, which are used to create an electronic version of the exam.

Essay analysis

Whether an essay is in draft-mode or in review, Turnitin AI flags areas for improvement and guides students through feedback. Essay analysis is performed with a suite of deep learning models, from BERT-based language models to CNNs, with advanced NLP tooling and statistical modeling.