Yandex B2B Tech has launched AI-SPM, a solution for managing the security and risks of AI systems, on the Russian market. It allows identifying potential vulnerabilities in models, data, and computing infrastructure, as well as controlling dangerous settings. Yandex B2B Tech announced the product launch, and Yandex Cloud CEO Grigory Atrepyev spoke about its capabilities. According to him, AI-SPM is among the first comprehensive solutions of this class on the Russian market.
The system collects the main elements of the AI infrastructure — models, data, and servers — into a single picture. Then it links them with identified vulnerabilities and suggests protection options. This allows to foresee potential problems and determine what to do with them.
For example, AI-SPM can help detect the risk of data spoofing used to train a model, the possibility of an attack through a specially crafted query, or excessive access by outsiders to the AI system. The tool also constantly monitors the infrastructure and searches for dangerous settings without the need for manual audits every time.
The solution became part of the Yandex Security Deck service. AI-SPM is based on the AI-SAFE 2.0 methodology, developed by Yandex Cloud and SolidLab. It is designed to assess AI security at different stages of its life cycle.
The market for such solutions in Russia is still forming. According to Yandex B2B Tech, its volume could reach 1 to 4 billion rubles in 2026, and grow to 11–30 billion rubles by 2029–2030. This is the company's estimate, not a confirmed independent market volume.