Softline Launches AI Security Services Practice
Softline Group has launched a new business unit dedicated to securing artificial intelligence systems. The practice is built on the expertise of Infosecurity Softline Solutions.
The company now offers AI security maturity assessments, risk analysis, Red Teaming for LLM applications, implementation of LLMSecOps practices, development of enterprise AI governance platforms, and support for organizations preparing for ISO/IEC 42001 certification.
Managing AI Adoption
The move expands the portfolio of one of Russia's largest IT integrators into the rapidly growing AI security market. It is a notable development as organizations increasingly deploy AI across software development, analytics, and customer-facing services. In parallel, they are confronting new risks, including data leakage through AI services, shadow AI use by employees, and vulnerabilities in LLM-based applications.
The trend could indirectly strengthen the security of banking platforms, telecommunications services, government systems, healthcare, and education environments where AI is increasingly deployed. For businesses and public-sector organizations, it represents a step toward more controlled AI adoption by helping determine whether new AI solutions could introduce data leaks or unreliable decision-making.
Softline's new practice also aligns with a broader global trend in which AI security is emerging as a dedicated cybersecurity discipline rather than simply another component of a conventional IT audit.
Toward Governed AI Operations
The launch could help accelerate the formation of a dedicated market for AI security services. As generative AI, LLM applications, AI assistants, and intelligent services become more deeply embedded across banking, retail, manufacturing, government, and software development, demand for specialized security services is likely to increase, especially as Russia's information security market continues to expand and regulators place greater attention on AI-related risks.
Export opportunities for the practice remain limited because of intense international competition and stringent foreign regulatory requirements governing trustworthy AI. The outlook in the domestic market, however, appears considerably stronger. Organizations increasingly need more than one-time model assessments. They require continuous AI risk management frameworks covering everything from employee policies governing generative AI use to regular testing of LLM applications for data leakage, prompt injection attacks, and guardrail bypass techniques. Those requirements are pushing the market beyond AI experimentation and toward production-scale deployment.
Growing Demand for Independent Risk Assessment
Russia has been developing AI governance frameworks and responsible AI practices since 2021. The process began with the adoption of the AI Code of Ethics, which established core principles covering trust, accountability, and the protection of human rights. That was followed by the National AI Development Center (Natsionalny tsentr razvitiya iskusstvennogo intellekta - National AI Development Center, NCDAI) and its partners launching an AI readiness index for different industries, where security is treated as a key indicator of mature AI adoption. During 2024 and 2025, Rosstandart introduced a series of AI standards covering everything from terminology to implementation requirements. Together, these measures laid the groundwork for more formal AI auditing and governance.
AI security became a much more prominent topic during 2025. Cybersecurity conferences increasingly shifted their focus away from AI's benefits toward the risks it poses to corporate operations and data protection. Russia's Federal Service for Technical and Export Control (FSTEC) went further by explicitly identifying AI technologies as a source of cyber threats in its official threat database, including attacks targeting models, datasets, and retrieval-augmented generation (RAG) systems.
Against that backdrop, one statistic stands out: nearly half of Russian companies deploying AI still do not allocate a dedicated budget for AI security, while only one-quarter have adopted formal governance policies. That gap is creating growing demand for external risk assessments, Red Teaming engagements, and the implementation of comprehensive LLMSecOps frameworks.
Scaling the Market
Softline's new practice highlights the Russian market's transition from large-scale AI deployment toward ensuring the reliability and cybersecurity of AI systems. For the company, it represents an opportunity to establish a strong position at the intersection of AI consulting, cybersecurity, and risk management.
Industry forecasts suggest that AI security services could soon become a standard component of large-scale AI deployment projects across finance, manufacturing, telecommunications, and the public sector. Experts expect the strongest demand to center on Red Teaming for LLM applications, assessments of data leakage risks, governance of shadow AI use by employees, and the development of internal policies for secure AI adoption.
