ITMO Humanities Students Participate in AI Thesis Initiative

42 students and 22 lecturers in social sciences and humanities from ITMO took part in the project. Students defended theses in which they used AI – legally and with permission from the university. Participating universities received a grant from Yandex Cloud for the use of YandexGPT in education, though students could choose to use any generative tools. Also within the project, experts from Yandex and Higher School of Economics (HSE University) taught lecturers about the use of AI tools – from the fundamentals of prompt-writing to creating one’s own AI assistant.

“The advance of AI radically changes education – both the academic process and the methods of producing knowledge, including graduation papers. And while AI is a common tool in engineering sciences, its applications in humanities are still developing. This project by Yandex and HSE University was of particular interest for the Faculty of Technological Management and Innovations because our students and heads of programs work with a great deal of texts, complex analytics, and interdisciplinary connections. As a result, we demonstrated that AI amplifies thinking in humanities, not replaces it. Our graduates spent less time on routine tasks and more time on creative solutions, which is a crucial skill for future managers and innovators,” says Andrey Anfinogenov, the dean of ITMO’s Faculty of Technological Management and Innovations.

This project is part of ITMO’s systemic efforts to integrate AI into education. At the university, AI is studied within specialized Bachelor’s and Master’s programs and actively applied in various fields, such as chemistry, genomics, biotechnologies, robotics, and urban studies. Moreover, the university employs an analytical monitoring system for the quality of admissions and education, as well as a system for automatic generation of lecture summaries from video.

“Here at ITMO, we are open to technologies and experiments. That’s why we always encourage new thesis formats; for example, digital humanities projects or startups. This diversity offers interesting options and trajectories for students and educational programs. And even more opportunities open up if we add AI into the mix. During this joint project with Yandex, we didn’t experience any administrative barriers or conservative backlash – there was only a lively discussion on how best to integrate AI into the academic process,” shares Aziz Ashirov, a PhD student of the Philosophical Anthropology program and a lecturer and development programs coordinator at ITMO’s DH Center.

At their defenses, students described the tasks they solved with AI services and their plans for applying them in their further work, while the project’s committee evaluated the students’ ability to critically assess the work of AI.

“Over the past year, AI has become a full-scale companion in solving educational problems; the job market now requires graduates to possess AI-related competencies. At Yandex Education, we have set ourselves an ambitious goal: to help universities develop AI literacy and provide students and lecturers with in-demand skills. This year, the AI thesis project involved several hundred students at over a dozen universities. It’s of particular note that students treat AI tools as partners, not as their replacement. We aim to continue helping universities find the optimal ways to integrate modern technologies,” says Kirill Barannikov, the head of strategic development of higher education at Yandex Education.

A survey of participating students demonstrates that they primarily used AI to structure texts (61%) and review literature (51%). 47% of students turned to AI for drafts and writing texts, 45% – for automation of data processing, and 43% – for text editing.

At the same time, students treat AI as a helpful tool that adds to their work, but does not substitute for it: 45% of students assessed AI’s contributions as moderate (the services helped them in some aspects), 34% – as insignificant (they used GPT sporadically and this didn’t affect the outcome of their work), 17% turned to AI during the entire workflow, while 4% stated that their participation in the project didn’t have an effect on their work (i.e. they hardly used neural networks). 

The project involved a total of over 500 students of humanities and social sciences, economics and  management, history and communications, philology, media, and pedagogy. Last year, 20 HSE University students joined the pilot launch of the project; in 2025, the list of participating universities includes ITMO, Kazan Federal University, Novosibirsk State Technical University, Lobachevsky State University of Nizhny Novgorod, Siberian State Medical University, North-Caucasus Federal University, Tomsk State University, Tyumen State University, Moscow Metropolitan Governance Yury Luzhkov University, and Ural Federal University.

YandexGPT is a neural network that can generate new texts and work with existing ones, suggest ideas, and consider the context of a conversation. The YandexGPT 5 family includes a powerful YandexGPT 5 Pro model, a beta version of the thinking-mode model, and a Lite version available as an open-source solution. The models are integrated into the Alice assistant and can help with various tasks, from personal to professional and educational; the models are also available in Alice’s “Let’s Think” mode. Moreover, YandexGPT summarizes texts found online via Yandex Browser or 300.ya.ru, as well as product reviews in Search and Yandex Market; the model also helps sellers generate descriptions of their products. Business owners can test the model in their work by applying for API access.