Xsolla Employees Fired by CEO Allegedly via Big Data Analytics of Work Activities

Xsolla CEO fired more than a hundred employees from his company in Perm, Russia, based on big data analysis of their remote digitized-work activity, which critics said was violating employee's privacy, outdated, and extremely ineffective.

Risk classification

• Primary risk domain: 6 Socioeconomic & Environmental
• Primary risk subdomain: 6.2 Increased inequality and decline in employment quality

The deployment of secret AI surveillance to track and terminate employees based on software activity metrics directly degrades the quality of employment and worker trust.

Additional risk subdomains

• 2.1 Compromise of privacy by obtaining, leaking or correctly inferring sensitive information: The AI system secretly tracked and analyzed employees' private activities and communications across workspace software without their knowledge.
• 7.3 Lack of capability or robustness: The tracking system failed to reliably measure actual productivity, relying instead on flawed proxies like network time and software interaction.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The risk of mass layoffs and secret surveillance was caused by intentional decisions made by Xsolla's management using the tracking system.

EU AI Act risk tier

• Risk tier: 2 High Risk

High Risk: The system is an employment and worker management system used to monitor and evaluate worker productivity, which is explicitly classified as high risk under the EU AI Act.

AI system and alleged parties

• AI system: Xsolla big data analysis AI
• AI purpose: Workforce Monitoring and Evaluation; Automatic Skill Assessment
• Behaviour type: Assistant
• Alleged developer: unknown
• Alleged deployer: Xsolla
• Alleged harmed parties: Xsolla employees

Harm severity

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Minor, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Minor, indirect Negligible
• Civil rights: direct Minor, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Minor, indirect Minor
• Psychological: direct Minor, indirect Minor
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Financial

Reported: Yes, 150 employees lost their jobs, resulting in immediate loss of salary, though some severance/compensation was mentioned.

Directly caused: 150 employees suffered immediate loss of employment and regular income due to the AI-driven termination.

Indirectly caused: N/A

Inferred additional harm: The 150 terminated employees likely faced ongoing financial instability and loss of livelihood while searching for new employment, despite receiving some severance pay.

Differential treatment

Reported: Yes, the reports note that the AI evaluation could unfairly target and harm those with nonstandard work approaches or different neurodiversities.

Directly caused: The AI system singled out and tagged 150 specific employees as unengaged based on software activity metrics, leading to their termination.

Indirectly caused: N/A

Inferred additional harm: The metric-based tracking likely disproportionately penalized employees with non-standard working patterns, remote workers, or neurodivergent individuals whose productivity is not accurately captured by active software time.

Civil rights

Reported: Yes, the reports mention concerns regarding employee data collection, privacy, and labor rights under Russian employment law.

Directly caused: The secret tracking of employees' activities across Gmail, Jira, and Confluence without their explicit knowledge violated their expectation of workplace privacy.

Indirectly caused: N/A

Inferred additional harm: Potential violations of local labor laws and data privacy regulations regarding the unauthorized surveillance of employees, especially those working from private home computers.

Privacy

Reported: Yes, the reports explicitly highlight concerns about data collection, secret tracking, and spying on employees' private activities.

Directly caused: Xsolla's big data team tracked employees' activities in Gmail, Jira, Confluence, chats, and documents without many employees being aware they were being monitored.

Indirectly caused: The shift to remote work during the pandemic meant this tracking potentially occurred on employees' private computers and home environments.

Inferred additional harm: The surveillance likely captured sensitive personal communications and activity data from remote workers' private devices without proper consent or disclosure.

Psychological

Reported: Yes, the reports state that tracking and terminating employees based on these metrics is sure to create distrust and anxiety within the workforce.

Directly caused: The sudden layoffs and public calling out of employees as 'unengaged' caused shock and distress among the 150 affected workers.

Indirectly caused: The remaining workforce experienced increased anxiety, stress, and distrust due to the realization that their activities were being secretly monitored.

Inferred additional harm: It is highly likely that many of the 150 terminated employees suffered prolonged stress, anxiety, and loss of self-esteem due to the sudden and public nature of their firing.

People affected

• Occurrences reported: 1
• People reportedly harmed: 150
• People reportedly exposed: 150

Potential causes

Management

• Growth Prioritization: Layoffs were driven by the company failing to sustain 40 percent growth.
• Lack of Algorithmic Transparency: Management failed to inform employees about tracking systems and metrics.

Technology

• Flawed Productivity Metrics: Algorithm evaluated software interaction rather than actual work output.
• Inability to Track Creative Work: AI cannot measure unconventional or neurodiverse creative work styles.

Data Inputs

• Secret Workspace Tracking: Employees were unaware that their software activities were being tracked.
• Proxy Performance Data: Using Jira, Confluence, and Gmail activity as a proxy for engagement.
• Remote Work Data Complications: Tracking remote employees on private computers raised privacy concerns.

Human Factors

• Lack of Leadership Empathy: CEO delivered insensitive termination emails and public social media posts.
• Workforce Distrust and Anxiety: Secret surveillance created a culture of fear and low morale.

Process and Methods

• Algorithmic Decision Making: Laying off 150 workers based solely on automated big data categorization.
• Ineffective KPI Design: Measuring network time encourages gaming metrics instead of productive work.

Regulatory Environment

• Gray Legal Area on Data Privacy: Lack of clear regulations regarding tracking remote employees on private PCs.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent details regarding the number of employees fired (150), the location (Perm, Russia), the date of the email leak (August 3, 2021), and the specific software tracked (Jira, Confluence, Gmail, etc.). The CEO's own statements and emails are quoted directly, confirming the use of big data/AI tracking for the layoffs.
• Ambiguities identified: The exact algorithm or specific AI model used by the big data team is not named, and the precise metrics used to define 'unengaged' are not fully detailed.
• Alternative interpretations: Some might view this as a standard big data analytics query rather than an advanced AI system, though the reports repeatedly refer to it as an AI-based analysis and data collection AI.

Xsolla terminated 150 employees in Perm, Russia, using AI-driven big data analysis of workplace software activity. While the incident sparked significant backlash regarding worker privacy and algorithmic surveillance, its national security impact remains negligible to minor, confined entirely to corporate labor practices and workplace privacy concerns.

• Overall national security impact: Minor
• Response level: Moderate
• Scope: Single nation
• Primary target: Russia
• Alleged perpetrator: Aleksandr Agapitov

Threat characteristics

• Imminence: Long-term. Represents an ongoing strategic and ethical concern regarding workplace surveillance rather than an active national security crisis.
• Autonomy: Human-supervised. The big data AI system analyzed activity and tagged employees, but human management made the final decision to execute the terminations.
• Novelty: Evolved capability. Represents a significant escalation in the scale and direct application of automated worker surveillance to execute mass layoffs.

Impact by dimension

• Physical security: Negligible. No physical systems, critical infrastructure, or kinetic threat capabilities were involved or affected by this corporate HR incident.
• Information security: Negligible. The incident involved internal corporate tracking of workplace software and does not represent an intelligence compromise or information warfare operation.
• Sovereignty: Negligible. This was a private commercial decision regarding company employees and had no impact on state authority or government functions.
• Economic security: Negligible. While 150 employees lost their jobs, this commercial layoff does not threaten strategic industries, national economic stability, or critical supply chains.
• Societal stability: Minor. The incident highlights concerns regarding workplace privacy, unauthorized employee surveillance, and the potential for algorithmic bias in labor decisions, but impact is limited to a single company.