AI-Powered Email Gateway: Solar Group and SEG-T Build a New Defense Against Phishing
Solar Group and the SEG-T team have begun developing a Russian secure email gateway (SEG) powered by a multi-agent AI architecture. The platform will be available as a cloud service, for on-premises deployments, and in Kubernetes environments, with deployment expected to take as little as 15 minutes.
The solution analyzes not only the technical characteristics of emails, links, and attachments but also their content. Its AI agents detect signs of social engineering, including urgency, fear, and psychological pressure, even when messages contain no malicious code. The system also recommends filtering rules, explains why messages were flagged, and generates attack reports for cybersecurity teams.
Security Tools Gain Ground
The joint investment and technology project aims to create a next-generation Russian secure email gateway for phishing protection using generative AI. The initiative expands the domestic market for enterprise email security products while supporting import substitution efforts. At the same time, it enters a competitive landscape that already includes BI.ZONE, F6, Kaspersky, UserGate, and Positive Technologies.
Strengthening email protection for businesses and government organizations should reduce the risks of data breaches and fraud. For Russia, that means a broader selection of domestic cybersecurity products and lower dependence on foreign technologies. Globally, meanwhile, the project reflects the broader trend of applying AI to cybersecurity. Its international commercial potential is still too early to assess and will become clearer only after the product reaches the market.
Why Email Content Matters
SEG-T is expected to find its primary applications in Russian government agencies, industry, financial services, telecommunications, healthcare, and IT, where email-based attacks remain a major threat. According to Positive Technologies, social engineering accounted for 50% of successful cyberattacks during the first half of 2025. Demand is also being driven by the continued growth of phishing: the number of such attacks increased by 33% year over year in 2024 and by 72% compared with 2022. The threat landscape continues to evolve. According to BI.ZONE, the volume of phishing emails targeting corporate systems tripled in 2025. Most attacks now originate from compromised user accounts rather than obviously malicious senders, increasing the importance of analyzing email content itself.
SEG-T's prospects will depend on how accurately it detects social engineering, how effectively it minimizes false positives, how well it handles Russian-language content and industry-specific terminology, and how seamlessly it integrates with email platforms and cybersecurity systems. Certification, successful deployments, and pricing will also play important roles.
The product's export potential is primarily tied to CIS markets. Native Russian-language support and rapid deployment could become competitive advantages. Expanding beyond those markets, however, will require adaptation to additional languages, regulatory frameworks, and national email platforms.
The Evolution of Multi-Agent Email Protection
Import substitution in cybersecurity accelerated between 2022 and 2024. After foreign vendors exited the market, organizations began migrating to domestically developed email gateways and security platforms. By 2025, the Russian SEG market already included products from BI.ZONE, F6, Kaspersky, Solar, and other vendors. At the same time, attackers adopted increasingly sophisticated techniques to bypass conventional defenses, including campaigns using encrypted archives and download links, making it necessary to analyze the entire interaction chain associated with each email.
During 2024 and 2025, developers strengthened SEG technologies. Against that backdrop, Positive Technologies introduced PT Email Security. Phishing attacks continued to increase in 2025 as threat actors made greater use of automation and personalization, reducing the effectiveness of template-based filtering.
By 2025 and 2026, the industry's focus had shifted from verifying email senders to analyzing message content, as attackers increasingly relied on compromised legitimate accounts. That shift accelerated the development of multi-agent protection models in 2026, including the SEG-T project, which combines distributed email inspection with automated security management.
SEG-T represents a Russian effort at the intersection of enterprise email security and multi-agent AI. Its relevance stems from the growing prevalence of phishing campaigns that rely on manipulation and legitimate compromised accounts rather than malicious files. If successfully commercialized, the platform could establish a position in the midmarket and large-enterprise segments by offering on-premises deployment, integration with Russian cybersecurity ecosystems, and lower administrative workloads. Its rapid deployment also makes it accessible to midsize organizations. Even so, the advantages of its multi-agent architecture have not yet been validated through independent real-world deployments. Its competitive position will become clearer after broader testing and production implementations.
