2023 • Bill No. LXIX of 2023
Artificial Intelligence (Protection of Rights of Employees) Act, 2023
Protecting employee rights in AI implementation at workplaces, ensuring transparency, fairness, and proper consent mechanisms for AI deployment affecting workers.
Overview
The Artificial Intelligence (Protection of Rights of Employees) Act, 2023, addresses the growing concerns about AI’s impact on the workforce. With 74% of Indian employees worried about losing their jobs to AI according to Microsoft research, this legislation provides crucial protections for worker rights in an AI-driven workplace.
Key Objective
To protect employee rights in relation to AI use and implementation in workplaces, regulate AI deployment, and ensure transparency in AI technologies that affect workers’ jobs, rights, and opportunities.
Why This Act is Essential
- Job Security Concerns: Indian employees are more concerned about AI job displacement than counterparts in US, UK, and Germany
- Algorithmic Bias: AI systems can perpetuate biases in hiring, promotions, and performance evaluations
- Privacy Issues: Increased data collection from workers with little transparency
- Autonomy Protection: Need for worker control over AI-driven decisions affecting their employment
Core Employee Rights
The Act establishes fundamental rights that all employees enjoy when AI systems are deployed in their workplace.
Right to Informed Consent
Right to Refuse AI Decisions
Right to Review Decisions
Right to Training
Right to Data Privacy
Right to Opt-Out
Consent Mechanisms
Section 4 establishes strict consent requirements for AI implementation in workplaces, ensuring employees have control over how AI affects their work environment.
Explicit Consent Requirements
Employers must obtain written consent from employees before implementing AI systems that:
- Job Security Concerns: Indian employees are more concerned about AI job displacement than counterparts in US, UK, and Germany
- Algorithmic Bias: AI systems can perpetuate biases in hiring, promotions, and performance evaluations
- Privacy Issues: Increased data collection from workers with little transparency
- Autonomy Protection: Need for worker control over AI-driven decisions affecting their employment
⚠️ Consent Best Practices for Employers
- Use plain language in consent documentation
- Provide specific details about AI system capabilities and limitations
- Explain data usage, storage, and sharing practices
- Offer genuine opt-out alternatives without penalty
- Allow consent withdrawal at any time
- Maintain comprehensive consent records
Equality Impact Assessment
Section 5 mandates comprehensive equality impact assessments to ensure AI implementations do not discriminate against employees based on protected characteristics.
Assessment Requirements
- Frequency: Every five years, with first assessment due five years after Act implementation
- Scope: Holistic evaluation of AI technology impact on individuals and groups
- Focus Areas: Race, gender, ethnicity, and other protected characteristics
- Responsibility: Appropriate government designs and conducts assessments
Assessment Outcomes
The equality impact assessment results guide policy adjustments, employer requirements, and enforcement priorities to ensure AI systems promote rather than hinder workplace equality.
Algorithmic Bias Prevention
The Act recognizes algorithmic bias as a critical concern, requiring proactive measures to detect, prevent, and remediate biased AI systems in workplace contexts.
Mandatory Prevention Measures
- Regular Auditing: Systematic evaluation of AI systems for biased outcomes
- Dataset Diversity: Ensuring training data represents diverse employee populations
- Bias Mitigation Techniques: Implementation of technical solutions to reduce algorithmic bias
- Continuous Monitoring: Ongoing assessment of AI system fairness in real-world deployment
- Corrective Action: Prompt remediation when bias is detected