Turnitin unveils AI paraphrasing detection for academic integrity

Turnitin has introduced a new AI paraphrasing detection feature, expanding its suite of artificial intelligence writing capabilities.

The feature aims to assist educators and publishers in identifying instances where AI paraphrasing tools have been used to modify AI-generated text, thereby addressing growing challenges in maintaining academic integrity in the face of proliferating generative AI tools.

AI-generated content presents unique difficulties in ensuring originality in education and research writing. Writers, including students and researchers, sometimes employ AI paraphrasing to alter AI-generated text, creating the false impression that such works are original. According to a study conducted by Tyton Partners, approximately 59% of students are regular users of generative AI tools, compared to nearly 40% of educators and administrators.

The new AI paraphrasing detection feature seeks to mitigate this issue. Annie Chechitelli, Turnitin’s Chief Product Officer, commented, “AI has incredible potential to benefit education, but writers need to be transparent in how they use Generative AI.

"Applying AI paraphrasing to AI-generated content is likely a sign that the writer is trying to hide their use of artificial intelligence, which is the opposite of transparency. Turnitin provides the tools and insights to preserve the value of original work," she said.

This initiative follows Turnitin's 2023 launch of its AI writing detection feature. Since implementation, over 200 million papers have been reviewed, revealing that approximately 11% contained at least 20% AI writing, and about 3% had at least 80% AI writing. The AI writing detection functionality forms part of Turnitin Originality and is accessible to iThenticate users via separate licensing.

By automatically checking each submission for both AI writing and AI paraphrasing, the feature highlights where a text spinner was possibly applied, offering insights into the text's construction. This new tool provides an overall percentage breakdown of documents likely to be AI-generated and those with AI-generated and potentially AI-paraphrased content.

Turnitin’s AI writing detection feature has been noted for its contribution to supporting academic integrity. It received recognition from the 2024 Bett Awards in the AI in Education category and was named Best in Show by Tech & Learning at ISTE 2023.

As artificial intelligence becomes an increasingly ubiquitous resource in academic settings, tools like Turnitin's AI paraphrasing detection are pivotal in fostering ethical writing practices. The feature not only aids in identifying AI-modified content but also encourages educators and researchers to engage writers in conversations about maintaining academic honesty and proper paraphrasing techniques.