Scientists at ITMO University in St. Petersburg have developed an artificial intelligence-based tool that determines with up to 94% accuracy whether a text was written by a human or an AI. The system is also capable of editing texts, reducing their "machine origin," and is already available in a demo version.
The new tool, created in ITMO's Computer Technologies Laboratory, analyzes the style and content of the text, identifying whether it was created by a human, AI, or paraphrased by AI. The algorithm successfully distinguished authorship in 94% of cases when tested on 5,500 Russian-language texts. For texts paraphrased by AI, the accuracy was 80%. The system uses two large language models that compare how "surprising" or "unexpected" the text is for them, and also analyzes linguistic features: word length, sentence structure, lexical diversity, and readability.
To train the classifier, scientists created a corpus of more than 4,000 texts in Russian, including scientific articles, essays, news, paraphrased texts, and materials generated by AI, such as ChatGPT and Gemini. An "obfuscator" was additionally developed - a tool that edits the text, eliminating traces of AI while preserving meaning and readability. It can be used to test the stability of detectors or prepare texts for publication.
A demo version of the tool is available on the Hugging Face Spaces platform, where any user can test their text. In the future, scientists plan to implement the service at ITMO to check student work and develop the project with the involvement of new researchers. The tool can be used in education, media, and business to label AI content and verify documents.
Read more materials on the topic:
The State Duma has defined the concept of "artificial intelligence"
Helping teachers!: "Znanie" has launched a course on how to detect AI in homework assignments
Now on home
The ship will be sunk in the Pacific Ocean, and its place will be taken by Progress MS-33
The company expects to continue operating even with mobile communication restrictions
Personalized devices transmit sound through the skull bones directly into the inner ear
Some UAVs have learned to hack directly in the air
A device with a lifting capacity of 500 kg will begin testing in two years
The enterprise will be able to assemble large-sized structures
The drone may have received AI and a mesh modem for navigation
The fleet has been updated, turn signals added, and insurance for pedestrians provided
Project 22350 ship practiced searching for and destroying submarines during exercises
Representatives of the carrier stated that there are no discussions with potential buyers
MEPhI Scientists Create Li₃C for Aviation - A Lightweight and Safe Material for Hydrogen Storage