Legislation to Safeguard Indian Workers from AI Threats, A New Bill in Parliament Sparks Debate
In recent years, artificial intelligence (AI) has been transforming industries worldwide, streamlining processes, improving efficiency, and reducing costs. However, this rapid adoption has also raised serious concerns over employment, workplace discrimination, and the erosion of workers’ rights. As India’s labour market, one of the most diverse and economically significant in the world, braces for the widespread adoption of AI technologies, a new private member’s bill has been introduced in the Rajya Sabha aiming to protect employees from the risks associated with AI in the workplace.
The bill, officially titled the Artificial Intelligence (Protection of Rights of Employees) Bill, has been proposed by Trinamool Congress MP Mausam Noor. Its primary focus is to safeguard workers’ rights in the age of AI, address algorithmic biases, and ensure fair treatment and transparency in automated decision-making. The bill, while narrower in scope than its European counterpart—the EU Artificial Intelligence Act—is deeply significant for the Indian context, where millions of jobs could be affected by automation.
Why This Legislation Is Timely
The push for AI-specific worker protection laws in India comes on the heels of high-profile layoffs in the information technology (IT) sector. Several job cuts have been linked—directly or indirectly—to the adoption of AI-based models by companies seeking efficiency gains. In a labour-intensive economy like India, where job creation remains one of the biggest challenges, the potential for mass job displacement due to AI is a pressing concern.
Globally, the risks posed by AI in hiring and workplace decision-making are already evident. For instance, Amazon’s AI-driven recruitment tool, introduced in 2014, was found to have a gender bias problem. After analysing a decade’s worth of resumes (which were predominantly from male applicants), the algorithm started favouring male candidates. It even downgraded resumes containing terms like “women’s chess club” or “women’s debate team,” inadvertently reinforcing gender inequality. Amazon ultimately scrapped the project. This case serves as a stark reminder of the unintended consequences of AI, especially when trained on biased historical data.
Key Provisions of the Bill
The Indian bill seeks to tackle such biases by mandating the “mitigation of biases in AI algorithms that may adversely affect employees’ rights or opportunities” through performance audits and impact assessments. Organizations would be required to regularly assess their AI systems to ensure fairness, non-discrimination, and compliance with established rules and regulations.
Some of the bill’s most notable provisions include:
1. Bias Mitigation and Transparency
-
Mandatory equality impact assessments every five years to review AI systems’ fairness, non-discrimination, and compliance.
-
Inspired by the EU’s approach, the bill encourages the registration of “high-risk” AI systems to enhance transparency and accountability.
2. Data Protection and Consent
-
Aligning with India’s Digital Personal Data Protection Act (DPDPA), the bill places explicit consent at the center of AI data use.
-
Employers would be required to obtain written consent from employees before processing their personal data through AI algorithms.
-
The bill raises a critical question: how to ensure consent is meaningful in a country as socioeconomically diverse as India, where employees may feel pressured to comply.
3. Employee Rights vs. Employer Powers
-
The DPDPA allows employers to process employee data for “legitimate uses” related to employment, such as preventing corporate espionage or safeguarding trade secrets.
-
However, the bill stresses employees’ rights to refuse certain AI-driven decisions, especially if these decisions impact their job security, promotions, or compensation.
4. Training and Skill Development
-
Recognizing the fear of job loss due to AI, the bill requires organizations implementing AI to invest in adequate training and upskilling opportunities for affected employees.
-
This aims to ensure workers can adapt and remain relevant in an AI-driven workplace.
Challenges in Implementation
While the bill is a step in the right direction, several challenges remain.
First, the concept of explicit consent in India is complex. In many work environments, the employer-employee power dynamic is unequal, making “voluntary” consent questionable. Workers may feel compelled to agree to data processing, fearing negative repercussions if they refuse.
Second, the requirement for training and upskilling could be seen as an additional cost by employers, particularly if their primary motivation for adopting AI is to reduce payroll expenses. For smaller firms, these requirements could be financially burdensome.
Third, given the rapid pace of AI development, the bill’s proposed five-year assessment intervals may be too infrequent. In technology terms, five years is an eternity—algorithms can evolve significantly in just a few months.
Finally, monitoring compliance will require robust enforcement mechanisms, which India’s regulatory framework will need to strengthen.
Balancing Innovation and Protection
The broader debate this bill sparks is about striking the right balance between innovation and protection. On one hand, AI offers tremendous opportunities for growth, efficiency, and competitiveness. On the other, without adequate safeguards, it risks deepening inequalities, entrenching biases, and displacing large numbers of workers.
For India, a country with a vast and varied workforce—ranging from highly skilled IT professionals to low-wage gig workers—any policy framework on AI must account for this diversity. Worker protections cannot be one-size-fits-all. They must be tailored to specific industries, job types, and levels of technological exposure.
Global Context
Internationally, AI legislation is gaining momentum:
-
European Union: The EU AI Act takes a risk-based approach, with stringent rules for “high-risk” AI applications, including those in recruitment, law enforcement, and critical infrastructure.
-
United States: While no federal AI law exists, several states have introduced bills addressing algorithmic discrimination in hiring.
-
China: The government has issued regulations for AI recommendation algorithms, including transparency requirements and restrictions on discriminatory practices.
India’s proposed bill is narrower in scope but significant in its focus on the employee-employer relationship, making it one of the first attempts globally to legislate specifically for AI’s impact on workers.
The Road Ahead
If passed, the bill would mark a critical step towards preparing India’s workforce for the AI era. However, it will need:
-
Clear definitions of what constitutes “high-risk” AI in employment.
-
Effective enforcement by bodies like the Data Protection Board and labour regulators.
-
Public awareness campaigns to educate employees about their rights under the law.
-
Periodic reviews to keep pace with AI’s rapid evolution.
Ultimately, the bill’s success will depend not just on the law itself, but on the willingness of all stakeholders—government, industry, and workers—to embrace a human-centric approach to AI adoption.
Five Key Questions and Answers
Q1: What is the main objective of the Artificial Intelligence (Protection of Rights of Employees) Bill?
A: The bill aims to safeguard employees’ rights in the face of increasing AI adoption in the workplace. It focuses on preventing algorithmic bias, ensuring transparency in AI systems, protecting personal data, and mandating training for employees affected by AI-driven changes.
Q2: How does the bill address the problem of AI bias?
A: The bill mandates organizations to conduct equality impact assessments every five years to evaluate AI systems for fairness and non-discrimination. It also requires measures to mitigate biases in algorithms that could adversely affect workers’ rights or opportunities.
Q3: What role does consent play in the proposed legislation?
A: The bill aligns with India’s Digital Personal Data Protection Act (DPDPA) by requiring employers to obtain explicit, written consent from employees before using their personal data in AI systems. This is intended to give employees greater control over how their data is used.
Q4: Why is the provision for training and upskilling significant?
A: AI is often adopted to reduce payroll costs, which can lead to job losses. By requiring organizations to provide training and upskilling, the bill seeks to help employees adapt to technological changes and remain employable.
Q5: What are the key challenges in implementing the bill?
A: Challenges include ensuring genuine consent in unequal employer-employee relationships, managing the financial burden on businesses for training programs, addressing the long five-year review cycle for AI systems, and ensuring effective enforcement of the law.
