Russian AI Research Ranks High in Global Science but Struggles with Commercialization
Russia has achieved a strong position in global artificial intelligence research, placing third among BRICS countries and twentieth worldwide by the number of papers presented at ten leading international conferences on machine learning and artificial intelligence.
The results come from a study published by the Scientometric Center of HSE University. Between 2020 and 2025, Russian organizations contributed 560 research papers that accumulated more than 12,300 citations. The average international citation rate stood at 3.59, exceeding India’s 3.10 despite a smaller overall volume of publications.
Russian scientific output demonstrates particular strength in the mathematics of machine learning, optimization, and formal concept analysis. Additional notable results appear in computer vision and speech technologies. More than 40 percent of domestic publications were prepared with business participation.
Among companies, Yandex leads in contributions. Among universities, HSE University and Skoltech are the top performers. Among non-profit organizations, AIRI stands out.
Despite these academic achievements, a significant gap persists between high-quality research papers and working commercial products. Development is hindered by shortages of computing resources, restricted access to high-quality datasets, and weak mechanisms for transferring research into market-ready solutions.
The lag is especially noticeable in natural language processing, AI agents, and infrastructure technologies. The Ministry of Digital Development has stated its intention to stimulate demand for domestic AI solutions, expand computing infrastructure, refine regulation, and speed up the adoption of scientific developments.
Related articles
redb.Identity Deploys Three-Layer Defense to Neutralize Stolen Access Tokens in OpenID Connect
redb.Identity, a custom OAuth 2.1 and OpenID Connect provider built on .NET, addresses the inherent risks of bearer access tokens by implementing three distinct protection layers. The first layer uses a Backend-for-Frontend pattern with Blazor Server and HttpOnly cookies to ensure no tokens ever reach the browser. The second layer applies DPoP proof-of-possession binding so that a stolen token remains useless without the corresponding private key. The third layer provides rapid revocation through RFC 7009 endpoints, refresh token rotation, idle timeouts, and a hybrid push-pull backchannel logout mechanism that guarantees no revocation is lost across replicas. The architecture also mitigates XSS vectors from direct application code, compromised npm dependencies, browser extensions, and localStorage. Additional routine controls include constant-time secret comparison, TOTP replay protection, rate limiting at multiple scopes, and sanitized proxy headers. The design explicitly acknowledges that BFF shifts rather than eliminates risk, requiring complementary CSRF protection via SameSite cookies and antiforgery tokens.
Russia Weighs Mandatory Use of Sovereign AI Models Like GigaChat in Government Sectors
Russian authorities are discussing a draft government resolution that would require state bodies and companies to prioritize fully domestic AI models in multiple regulated areas even when they cost up to 50 percent more than alternatives. Only GigaChat developed by Sber currently meets the proposed sovereign criteria of being created by a Russian legal entity and running exclusively on Russian infrastructure. National models such as Yandex’s Alice AI would be allowed to incorporate open-source components. The measure would apply to law-making, national projects, taxes, customs, critical information infrastructure, education, healthcare and Gosuslugi services. The Ministry of Digital Development stresses that the list remains preliminary and open to further input from market participants. Industry players including Yandex and MWS AI have voiced concerns over reduced choice and potential quality or cost disadvantages for end users.
GTA: Vice City Runs Natively in Browser via reVC Port, Supports Smartphones and High Frame Rates
A browser-based version of the 2002 classic GTA: Vice City has been made available on the Quenq website, allowing play without any client installation on both desktop computers and mobile devices. The port supports keyboard, gamepad, and touch controls while offering save management, mod support, cheat code entry, and frame rate customization. Testing on an OnePlus 12 smartphone showed smooth performance in both landscape and portrait orientations. The implementation relies on the reVC decompilation project combined with WebAssembly for C++ execution and WebGL 2 for graphics acceleration. On desktop hardware the game can exceed 100 FPS, far above the original 30 FPS target, which can affect physics behavior. The site also hosts a similar GTA III port based on re3, with traces of the earlier DOS Zone project visible in the pause menu. Previous DMCA action by Take-Two in 2025 temporarily shut down the service, yet the versions remain accessible despite ongoing copyright concerns.
Basis Acquires 70% of Proto to Integrate Observability Platform into Basis Dynamix
Russian infrastructure software developer Basis has acquired a 70% stake in Proto, the creator of an observability and operational data analytics platform. The transaction was executed through a capital investment, with funds allocated to product development and integration into Basis’s ecosystem. Proto will operate as a subsidiary while retaining its team and continuing to advance the Proto Observability Platform. The technology will be embedded into Basis Dynamix to enable predictive analytics, incident forecasting, and proactive alerts before infrastructure failures occur. AI-driven features from Proto will support predictive assessment of IT system health. Standalone configurations will remain available for customers outside the Basis ecosystem. The company projects combined revenue from Proto technologies within its portfolio to reach at least 1 billion rubles between 2026 and 2031. CEO David Martirosov noted that the acquisition shortened development timelines by approximately 18 months compared to building similar capabilities from scratch. Proto ranked among the top three Russian observability platform providers in the 2025 Technologies of Trust study, with the domestic market estimated at 3–15 billion rubles and growing at roughly 24% annually.