The security of neural networks is one of the key challenges for AI technologies. This was stated in an interview with the newspaper "Respublika" by the leading researcher of the laboratory of biomedical images analysis of the UIIP NAS of Belarus Vasily Alekseevich Kovalev. He explained why it is already important to protect AI systems from attacks that can distort medical data, mislead neural networks and affect the diagnosis of diseases. Scientists of the institute are already studying the types of attacks and developing protection methods - so that AI technologies serve people, and do not create new risks. Read more in the material of the newspaper "Respublika" Publishing house "Belarus Segodnya"