HabrAugust 21, 2026🇷🇺Translated from Russian

Rospotrebnadzor and FAS to Extend Oversight to Websites: Automating Foreign Word Replacement Using LLM

Russian compliance specialists have created an automated system to help websites comply with Federal Law No. 168-FZ on protecting the Russian language ahead of expected enforcement actions by Rospotrebnadzor and the Federal Antimonopoly Service.

The law prohibits the use of foreign words when common Russian equivalents exist and bans both Latin script and transliterations written in Cyrillic. It requires identification of prohibited borrowings by comparing text against four official dictionaries approved by the Russian Academy of Sciences: orthographic, orthoepic, explanatory, and the dictionary of foreign words.

The developed microservice, named Normograph, implements a five-stage pipeline that significantly reduces the volume of text sent to the language model. First, HTML pages are cleaned of markup and tokenized. Second, primary filtering removes words present in normative dictionaries, user-defined white lists, and registered trademarks. Third, the system flags potential issues including words absent from dictionaries, known anglicisms, and Latin-script terms.

Only after these algorithmic filters does the LLM receive small relevant fragments. The model determines whether a word qualifies as an impermissible borrowing, considers context, and proposes accurate Russian synonyms that match in gender, number, and case. The pilot uses GigaChat API but the architecture allows easy substitution with other models such as YandexGPT.

An additional OCR module based on Yandex Cloud Vision extracts text from images collected by a crawler. The same multi-stage filter is applied to recognized text, with caching and size-based filtering to control costs. Logos and registered trademarks are automatically excluded from checks.

Testing showed that manual review of one page previously took about one hour, while the automated system reduces the process to between two and ten minutes. The solution is positioned as a corporate compliance tool for pharmaceutical, fintech, and large e-commerce platforms that must adapt hundreds of pages before regulatory inspections begin.

Related articles

HabrPolicy & Regulation

Costly Mistakes: How Russian Businesses Risk Millions in Fines for Personal Data Violations

A year after stricter Russian personal data protection fines took effect, many entrepreneurs continue to commit violations that could trigger multimillion-ruble penalties from Roskomnadzor. The article details ten common breaches, including the prohibited use of Google Forms for data collection, missing cookie banners, absent or invalid consent forms under forms, and failure to obtain separate consents for publishing reviews. Additional violations cover missing privacy policies, outdated notifications to Roskomnadzor, improper transfer of employee data to third parties without written consent, lack of data processing agreements, and absence of records for paper-based data storage locations. Each violation is explained with direct references to the Law on Personal Data, the Code of Administrative Offenses, and specific government orders, along with exact fine ranges for citizens, individual entrepreneurs, and legal entities. Practical remediation steps are provided, such as replacing foreign services with Yandex Forms, drafting compliant consent texts per Article 9, and submitting updated notifications under Order No. 180. The guidance emphasizes conducting a full site audit and implementing all required documents to avoid penalties throughout 2026.

HabrPolicy & Regulation

Rethinking SSO: Centralized User Data Provision and Authorization Processing in Corporate Systems

The article examines Single Sign-On systems not merely as authentication gateways but as architectural hubs for delivering user attributes and executing additional authorization logic. It highlights how SSO can aggregate data from sources like Active Directory, HR systems, and IDM platforms, then deliver it via OIDC claims to downstream applications. The discussion covers the shift from fragmented integrations across dozens of apps to a single trusted enforcement point using standards such as aggregated and distributed claims. It also explores the authorization pipeline where SSO acts as a Policy Enforcement Point querying external Policy Decision Points via the AuthZEN Authorization API 1.0. Practical examples include electronic business cards, role assignment, access routing, and mandatory MFA checks before token issuance. The piece stresses maintaining data ownership with source systems while establishing SSO as the single point of trust for applications.

AntiMalwarePolicy & Regulation

Bitrix24 Releases Fully On-Premise BI Constructor for Regulated Enterprises

Bitrix24 has introduced a new delivery model for its BI Constructor that allows complete deployment inside a customer's own infrastructure. The update eliminates any requirement for external servers, cloud APIs, or internet connectivity, ensuring that all corporate data remains within the organization's closed perimeter. Previously, even the boxed version of the platform needed access to external infrastructure for updates and auxiliary services, creating conflicts with internal security policies and regulatory demands in highly regulated sectors. The new on-premise variant performs all data processing and storage exclusively on customer servers, giving organizations full control over access rights, backups, updates, and integration with internal protection tools. The solution is compatible with the boxed edition of Bitrix24 running on PostgreSQL and does not connect to external CDNs or cloud services. Bitrix24 expects strong interest from large enterprises and organizations handling restricted-access data that must stay inside the corporate network. Pilot implementations have already been completed, with broader customer pilots planned in the coming months.

HabrPolicy & Regulation

Russia's Data Leak Penalties: 2.6 Million Rubles in Fines Despite 1.58 Billion Records Exposed in 2025

Russia introduced turnover-based fines for personal data leaks through Federal Law 420-FZ in late 2024, fundamentally altering the economics of information security investments. Over the first 18 months, Roskomnadzor opened 52 administrative investigations and issued 40 protocols totaling just 2.6 million rubles in penalties, with zero turnover fines applied. This occurred against a backdrop of 1.58 billion compromised records in 2025 alone. Public data leaks dropped fourfold in the first half of 2026, yet trading activity on underground forums rose nearly 60 percent as operators shifted to private sales. The law now ties penalties directly to the number of affected individuals and adds a turnover component for repeat violations under Article 13.11 of the Code of Administrative Offenses. Analysts note that the mere threat of larger fines has prompted companies to reassess data retention policies and risk models even without actual enforcement precedents.