AI in the Russian Workplace

Labour Law Framework for AI-Driven HR Decisions

Russian labour law, codified in the Labour Code of the Russian Federation (Trudovoy Kodeks Rossiyskoy Federatsii, TK RF), does not contain provisions specifically addressing artificial intelligence in the workplace. General principles of employment law apply, creating both opportunities and legal risks for employers deploying AI in human resources functions. The constitutional right to work (Article 37 of the Constitution) and the principle of non-discrimination in employment (Article 3 TK RF) form the baseline against which AI-driven HR decisions are assessed. Automated recruitment screening, performance evaluation, and termination recommendations must comply with the prohibition on discrimination based on gender, race, nationality, language, origin, property status, and other grounds. Where an AI system produces outcomes that disproportionately affect protected categories, the employer may face liability under Article 5.62 of the Code of Administrative Offences (KoAP RF), which imposes fines of up to 100,000 RUB for discrimination in labour relations.

The Federal Law on Personal Data (152-FZ) imposes additional constraints on AI processing of employee data. Article 86 TK RF requires employers to obtain employee consent for the collection and processing of personal data, except where processing is necessary for compliance with labour legislation or for the performance of the employment contract. Automated decision-making based solely on algorithmic processing that produces legal consequences for the employee raises questions under Article 16 of 152-FZ, which prohibits decisions based solely on automated processing of personal data, except with the data subject’s written consent or as otherwise provided by law. Employers deploying AI for hiring, promotion, or dismissal decisions should implement human review mechanisms to mitigate this risk.

Employee Monitoring and Surveillance

Russian law permits broad employee monitoring, subject to certain procedural requirements. Article 21 of the Labour Code recognises the employee’s right to privacy, but this right is not absolute and must be balanced against the employer’s right to control the production process under Article 22 TK RF, which includes the right to require employees to comply with labour discipline. The Constitutional Court of the Russian Federation has held, in Ruling No. 21-P of 2014, that employers may monitor employees’ use of work equipment and communications, provided employees are notified of the monitoring in advance and the monitoring is justified by legitimate business purposes.

Employers commonly deploy keylogging software, screen recording, email monitoring, and video surveillance in the workplace. The legal requirements for such monitoring include: (1) notification in the employment contract or internal labour regulations (pravila vnutrennego trudovogo rasporyadka); (2) a legitimate purpose, such as protection of trade secrets or compliance with safety requirements; and (3) proportionality — the monitoring must not intrude beyond what is necessary. Biometric monitoring, including facial recognition for access control and productivity tracking, is regulated by Federal Law No. 572-FZ of 2023, which requires written consent for biometric data processing and notification of Roskomnadzor. The Federal Service for Supervision of Communications, Information Technology and Mass Media (Roskomnadzor) has issued guidance requiring that biometric data be stored separately from other personal data and protected using FSB-approved cryptographic means.

AI Training Data and Employee Privacy

The training of AI systems on employee-generated data raises complex legal questions under Russian data protection law. Employee emails, chat messages, productivity metrics, and behavioural data may constitute personal data within the meaning of 152-FZ, particularly where the data can be linked to identifiable individuals. The principle of purpose limitation (Article 5, 152-FZ) requires that data collected for employment purposes be used only for the purposes for which it was collected, unless separate consent is obtained. Repurposing employee data for AI model training without explicit consent is likely to violate this principle.

The data localisation requirement under Article 18(5) of 152-FZ requires that databases containing personal data of Russian citizens be physically located in Russia. This applies to AI training datasets that include employee personal data. Cloud-based AI services that process employee data on servers outside Russia may violate this requirement, exposing the employer to administrative liability and potential blocking of the relevant service by Roskomnadzor.

Automated HR Systems and Digital Ruble Integration

The Russian Central Bank’s digital ruble (tsifrovoy rubl), introduced under Federal Law No. 340-FZ of 2023, creates new possibilities for AI-driven payroll and compensation systems. The digital ruble platform uses distributed ledger technology and supports programmable payments, which could enable automated wage calculations, tax withholdings, and benefit distributions through smart contracts. The Bank of Russia has published a digital ruble concept paper suggesting that automated payroll systems using the digital ruble could reduce transaction costs and improve transparency in employer-employee financial relations. However, Article 136 TK RF requires that wages be paid in Russian roubles and that the method of payment be specified in the employment contract, meaning employers cannot unilaterally introduce digital ruble payments without employee agreement.

The interaction between AI-driven HR decision-making and the digital ruble also raises questions under the Federal Law on the National Payment System (161-FZ). Automated salary calculations and deductions must comply with Article 5 of that Law, which requires the payer’s express consent for each payment operation.

Liability for AI-Caused Harm in Employment

Employers deploying AI systems in the workplace face potential civil liability for harm caused by algorithmic decisions. Article 1068 of the Civil Code (GK RF) holds employers liable for harm caused by their employees in the course of employment. Where an AI system wrongfully denies a promotion, terminates employment, or makes an erroneous compensation calculation, the employer may be held liable for material damage (materialny ushcherb) and moral harm (moralny vred) under Articles 232 and 237 TK RF. The employer’s liability may be difficult to disclaim through contractual provisions, as Article 9 TK RF renders void any employment contract terms that limit employee rights or reduce guarantees established by labour legislation.

The Code of Administrative Offences also provides for employer liability for violations of labour law arising from AI system failures. Article 5.27 KoAP RF imposes fines of up to 100,000 RUB for legal entities and up to 20,000 RUB for officials for violations of labour legislation, with higher penalties for repeat offences. Criminal liability under Article 145.1 of the Criminal Code (UK RF) applies where non-payment of wages, pensions, scholarships, or other benefits is committed out of selfish or other personal interest by the head of an organisation, with penalties of up to three years’ imprisonment.

Roskomnadzor Oversight and Enforcement

Roskomnadzor exercises regulatory oversight over AI systems that process personal data in the workplace. The Service has the power to conduct scheduled and unscheduled inspections, issue orders to remedy violations, and impose administrative penalties. Roskomnadzor’s 2024 guidance on AI and personal data states that AI systems used for HR purposes must undergo a data protection impact assessment (otsenka vozdeystviya na zashchitu dannykh) prior to deployment and must implement technical measures to ensure data accuracy, prevent bias, and enable individual review of automated decisions. Employers should maintain documentation of their AI systems’ technical specifications, training data, and validation results to demonstrate compliance during inspections.