AI Regulation in Science: Expectations of Scientists and Organizations (ISSEK HSE Express Information)
The Institute for Statistical Studies and Economics of Knowledge (ISSEK) at HSE University surveyed scientists and heads of research organizations to understand their expectations for state support of AI in science. Most respondents favor enabling conditions over strict controls, with over two-thirds supporting access to infrastructure, funding, and services. The findings were reported by CNews.
The Institute for Statistical Studies and Economics of Knowledge (ISSEK) at HSE University, based on surveys of scientists and heads of research organizations, studied what role they assign to the state in supporting AI in science and what kind of help they expect. The empirical base included two online surveys: one of heads of state universities and research institutes (May-June 2026, N=424 organizations, of which 250 universities and 174 research institutes), and another of Russian scientists (May 2026, N=1209 respondents, of which 769 work in universities, 318 in research institutes, and 122 combine work in both types of organizations). More than two-thirds of scientists (67%) and 72% of heads believe the state should primarily create favorable conditions for AI use in research. About 70% of scientists and 60% of heads do not support strict measures such as restrictions on AI use, target indicators, and monitoring. There is no consensus on acceptable forms of state regulation: 58% of scientists and 36% of heads support non-interference in technology choices, while about 40% in both groups favor more active state participation, including control. More than 80% positively assess measures like centralized access to computing power, targeted funding for equipment, support for interdisciplinary teams, payment for AI service subscriptions, long-term funding for AI projects, and creation of repositories for datasets and models. About 70% positively assess institutional measures like quality standards and recommendations, but less than half expect positive effects from a system to evaluate AI use efficiency, with 29% of scientists and 22% of heads thinking it would worsen research conditions. The surveys show the academic community expects infrastructure, funding, services, and data access, not strict control or directives. This supportive model aligns with approaches of scientific leader countries that have adopted national AI transformation strategies, and active AI use is a priority in Russia's Strategy for Scientific and Technological Development.
Source: CNews —
original
