Written by Saamragyi Gupta & Manya Gupta students at Jindal Global Law School, O.P. Jindal Global University.
Apps like Swiggy, Zomato, and Uber have formed an integral part of our everyday lives. It has replaced the lengthy market visits and made our daily-need products accessible anywhere and almost everywhere. These apps classify as ‘aggregators’ under Section 2(2) of the Code on Social Security, 2020 (“Code”), which engage gig workers, whom we see as delivery people at our doorstep. With a smooth change in consumer life, the gig economy has also transitioned outside of the traditional employer-employee relationship. Their activities are monitored by algorithms engaged by these aggregators, which then have a say in even deactivating accounts of these gig workers objectively based on parameters set by these aggregators in the coding of these algorithms. This is a troubling reality for the gig workers, as their livelihood depends on vague algorithmic ratings and customer reviews. These workers can be abruptly removed from the platform based on one negative rating, and are required to pay an additional deposit just to continue working. The issue is aggravated further due to a lack of an appeal mechanism and transparency surrounding the workings of these algorithms. The promise of flexibility for women workers often entails a trade-off, where speed and convenience are prioritised at the expense of their dignity, making them subject to harassment and dangerous working conditions.
The root cause of uncertainty faced by the gig workers is ambiguity in their status as employee or independent contractor. Even if they are considered to be employees, the next question is who is the employer when the control over them is exercised by algorithms. The Supreme Court is yet to decide in this regard in the case of The Indian Federation of App-Based Transport Workers (IFAT) v. Union of India, where one of the issues is whether gig workers can be classified as employees. While the decision on this issue is pending, the Parliament has gone a step ahead by officially recognising ‘gig workers’ as an altogether separate category from employees or independent contractors in the new labour codes. However, the bigger question on the protections provided to them under the labour regime still remains unanswered and cannot be understood unless their status as employee or independent contractor is determined. To answer this, when classical tests such as the control and supervision test and the integral business test are applied, there is a compelling case for recognising these gig workers as employees. The case of Dharangadhara Chemical Works v. State of Saurashtra introduced the control and supervision test, which relied on the principle that if the alleged employer has the right to control not just ‘what’ the person does but ‘how’ they do it, i.e. the manner, method, time, and place of work of the person, then an employment relationship exists. Some ride-hailing companies demonstrate this degree of control by monitoring the behaviour of these gig workers through algorithmic systems by using tactical notifications such as “complete 3 more rides to unlock a bonus”, which nudges a worker towards continued availability for work that serves the aggregator’s interests, but not necessarily the workers.
Further, in the case of HussainBhai v. Alath Factory Employees Union, the court laid down the integral business test to say that when a person or group of employees contributes labour to create goods or services for the benefit of another’s business, that other is the employer. The test is to check whether the person’s work is an integral part of the employer’s business, or merely accessory to it. In this regard, gig workers’ labour forms part of the very core commercial offering of the aggregator. Thus, the aggregators’ economic dependence on these gig workers all point towards classifying them as employees. Additionally, the courts in Dharangadhara Chemical Works and HussainBhai have urged the need to lift the veil and determine control by analysing who has actual power over the hiring and firing of the worker. Similarly, since the aggregator exercises control over the worker, it should be considered as the employer.
Despite this, a straightforward reclassification as ‘employees’ carries significant practical drawbacks that must be confronted. First,if every gig worker had to individually prove they were an employee in court through these tests, aggregators would challenge each case separately. For a worker who has just lost access to their account and their income cannot afford to fight a long legal battle. This is probably why the new labour codes use such a broad, catch-all definition of “gig worker” under Section 2(35) of the Code on Social Security, 2020. It has been deliberately made to include all kinds of working arrangements that don’t fit into the traditional employer-employee framework so that the workers don’t bear the burden of proving their status every time there is a dispute, especially considering that there is no clarity on their classification. Second, gig work is not a one-way formula that only benefits the aggregators. For many people, like migrant workers or those without formal education, it has become an important and accessible source of income. It even provides an additional income or part-time income for people. The inherent flexibility in gig work provides more opportunities and financial security to these workers. Bringing them under a traditional employment structure will defeat the very reason why the gig economy is booming. It might end up discouraging the aggregators from engaging gig workers in their business and leave the workers high and dry for employment.
B. The Real Problem: Algorithmic Control and Opacity
The central issue, therefore, is not whether gig workers should be classified as employees but whether they are adequately protected from the ambiguity surrounding the algorithms. It is often unclear how algorithms determine allocation of work, impose penalties, or lead to deactivation. In this sphere, the irony is that technology which is intended to make work more efficient has, in fact, made it more unpredictable and unclear. The current Code, while recognising gig workers as a distinct category, fails to provide safeguards to prevent their harassment arising from arbitrary, unchecked algorithmic decision-making. Given the unique position of gig work, there is a need for a tailored approach which, without undermining the flexibility that is the sector’s defining boon, also protects workers from unfair treatment.
This unique position of gig work becomes complex in determining accountability. They are often made to appear like independent contractors to further the aggregator’s interests. It is true that gig work does not entirely resemble traditional employment, but platforms exert substantial control over the workers with respect to the performance of their work and their continuance on the platform, thereby blurring the line between gig work and traditional employment. This goes beyond the level of control expected in an independent contractor relationship. This sui generis status justifies extending to gig workers termination protections available under the Industrial Disputes Act, 1947 (now the Industrial Relations Code, 2020), like that of prior notice of one month. Such a protection is necessary since when workers are deactivated based on algorithmic decisions alone, without being given an opportunity to be heard or a prior notice, it is equivalent to a termination and thus a call for similar safeguards is justified. It is pertinent to note, however, that procedural protections by itself are insufficient and must be coupled with substantive standards for deactivation that ensure fairness, transparency and a review mechanism of algorithmic decisions.
The International Labour Organisation (“ILO”), in its report on piece-rate work, has suggested a fixed methodology in calculating the wages and working hours of piece-rate workers. It recommends including a Personal Time, Fatigue, and Delay (“PF&D”) factor in the time and motion analysis of working hours. This standard can be imported and applied to gig workers as well, since the nature of their work is also task-driven, which is equivalent to a per-piece formula. This will account for the arbitrary standard set by algorithms, which does not factor in traffic density, adverse weather, demand surges, and permissible rest periods and penalises workers for circumstances beyond their control. Thus, requiring aggregators to incorporate factors like the location, time of day, customer demand, and unavoidable delays based on the PF&D formula into their algorithms will not sacrifice their governance of these workers; rather, ensure that the algorithmic decision-making is fair, accurate and reliable.
C. Reforming Algorithmic Governance in India
Comparative frameworks from various jurisdictions frequently provide helpful examples that India could incorporate into its labour regime. Spain’s Rider’s Law (2021) and the EU Directive (2024) both acknowledge an employee’s right to know about algorithmic systems that influence their working conditions. What is noteworthy here is the shift in burden of proof, i.e. the responsibility to explain how decisions are made falls upon the aggregator, not the worker, unlike India, where a gig worker’s account can be deactivated without ever being told the reason.
Spain has codified the right of the works council to be informed by the aggregator of the parameters, rules and instructions on which algorithms or artificial intelligence systems are based that affect decision-making that may affect working conditions, access to and maintenance of employment, including profiling. It has gone a step further by encouraging the aggregators to voluntarily collaborate with the workers’ trade unions in constituting a joint ‘algorithmic committee’ as a human-overseeing body over the algorithms used and has representatives from both workers and the aggregators. This has allowed collective bargaining through trade unions in Spain, which is entirely absent in India. So, the question now is, if a worker cannot understand the algorithms controlling them, and cannot collectively challenge them, what real recourse do they have? The honest answer, under current Indian Law, is none. However, a notable exception is the Karnataka Platform-Based Gig Workers (Social Security and Welfare) Act, 2025 (“Karnataka Act” or the “Act”), a state-level legislation specifically addressing gig worker rights. It mandates platform registration, a grievance redressal mechanism, a penalty upon aggregators for violations of the Act, as well as the constitution of a welfare board. While the recently notified Social Security (Central) Rules, 2026 (the “SS Rules”) also mandate similar registration of these gig workers on the e-Shram Portal and institute the National Social Security Board for gig workers, the Karnataka Act, in addition, mandates each aggregator to institute an Internal Dispute Redressal Committee (“IDRC”) which addresses the grievances of the gig workers and also provides appellate jurisdiction to the welfare board over decisions of IDRC unlike the SS Rules or the Code. However, this Act is not sufficient to conclude that gig workers across India are not exploited. It only gives workers a grievance forum but does not tell them what they are being accused of. Additionally, it fails to address the core issue of how work is allocated and controlled by algorithms.
The EU framework also introduces periodic impact assessments of algorithmic systems. A legitimate critique of algorithmic management is that its entire purpose is to reduce the need for human intervention, thus making human oversight mechanisms like those in the EU or Spain seem counterintuitive to the very logic platforms operate on. Yet this is precisely where the problem lies. When human judgment is removed from decision-making, accountability often disappears from such decisions. The need for this mechanism is significant for women workers since reporting harassment can risk a lower rating on the app and potentially result in their deactivation. Such problems cannot be cured by an algorithm that does not know the difference between a worker who performed poorly and a worker who refused to be mistreated. For a woman with no formal employment alternative, the choice between tolerating harassment and risking her income is not a choice at all. India could adopt the EU-based approach by mandating periodic impact reviews of platform algorithms, along with professional human oversight to override automated decisions where necessary.
EU and Spain model on algorithmic oversight provides a base framework that India can adopt while keeping in mind the social, economic, and technical factors specific to India. Therefore, we suggest that these platforms should be statutorily required to register their rating and deactivation algorithms with the relevant labour authority and disclose the criteria by which workers are evaluated and removed. This will ensure strict regulatory reporting and allow the relevant labour authority to scrutinise the criteria for deactivation. Additionally, this would also allow transparency in decision-making for the workers and contest arbitrary decisions. Further, certain amendments to the Code, like mandating a notice period before deactivation or creating an internal dispute redressal committee like under the Karnataka Act, can help strengthen the protections to these gig workers. Lastly, the adoption of ILO-aligned time and motion standards in algorithmic coding would ensure that technology does not become a means to bypass rights that Indian labour law has, over decades, carefully evolved to protect.
Gig work has transformed the way of living for many Indians because of its flexibility and inclusiveness. People who were earlier excluded from the job market have found a new way of earning, and that too from multiple sources simultaneously. However, this shift has also left workers exposed to several vulnerabilities. The debate about their classification as employees is not the issue now; rather, the focus is on resolving the uncertainty that surrounds the algorithm controlling them. Workers are rated, penalised, and deactivated by systems they can neither question nor challenge. Comparative models from Spain and the European Union could be incorporated in India to uphold the dignity of the workers along with the fairness of the system. It’s high time that these reforms are brought into the legal framework to create a right balance between ease of doing business and the right of fair working conditions. We align with what the opposition has brought to the table in the Parliament about 10-minute delivery and requiring the big sharks of the gig economy to hold accountability for the working conditions of these workers. There is hope that the National Social Security Board for the welfare of gig workers will raise and address the issue of algorithmic deactivation at the executive level in the meantime. We do see some changes being brought by these aggregators on their own. For e.g. Blinkit has started with on-the-go medical units for its workers at certain locations in Delhi. However, until Indian law requires aggregators to answer the basic questions on the algorithms they use, the promise of flexibility will continue to conceal a deeper insecurity.
Caveat: The views, analyses, and information presented in this article are provided in good faith and for general informational purposes only. No representation or warranty, express or implied, is made regarding the accuracy, adequacy, validity, reliability, or completeness of the information. Readers should conduct their own research and seek professional guidance where appropriate. Neither the author nor the publisher shall be held responsible for any loss, liability, or consequence arising from reliance on this content.
Leave a Reply