Sergei Kruglikov noted that adopting a legislative act is a lengthy process. “It can take several years. The model law 'On artificial intelligence technologies' was adopted by the CIS Interparliamentary Assembly on 18 April 2025. Its development began in May 2023. Initially, it was expected to be adopted only in November, but in fact, we did it ahead of schedule, which is a good thing. The law is recommended for use in national legislation by all CIS member states. Now that the foundation is in place, we need to develop a concept for AI development in Belarus. This will be a short-term document because technology evolves faster than people can comprehend it. It must be done at a high legislative level. The product must be verifiable in any subject area – only then will it be in demand and beneficial,” the institute director shared.

He added that large-scale AI implementation in Belarus requires a package of standards. “We hope that, as previously planned, a national technical committee for AI standardization will be established at the United Institute of Informatics Problems of the National Academy of Sciences of Belarus. Over 30 partners have already responded, willing to join the committee and provide experts to develop standards and other regulatory frameworks in this field,” Sergei Kruglikov said. “We must account for the risks posed by any technology, including AI. By properly understanding them, we can determine what legal acts are needed to minimize those risks. AI can be implemented everywhere, but who can confirm that what’s inside the 'black box' is correct? It must be verified and guaranteed that it will not cause harm.”

belta.by