Artificial Intelligence In E-Commerce - Consumer Manipulation And Product Liability Gaps In India’s Legal Framework

Author: LawSphereX

Category: Technology, Cyber & Privacy

Published: 2026-08-23T18:18:29.4562+00:00

Updated: 2026-08-26T17:12:01.785+00:00

ARTIFICIAL INTELLIGENCE IN E-COMMERCE:

CONSUMER MANIPULATION AND PRODUCT LIABILITY

GAPS IN INDIA’S LEGAL FRAMEWORK

Rudragouda M. Hommaradi, Research Scholar at University Law College & Department of Studies in Law, Bangalore University, Bengaluru Dr. Suresh V. Nadagoudar, Senior Professor & Research Supervisor at University Law College & Department of Studies in Law, Bangalore University, Bengaluru

ABSTRACT

The application of Artificial Intelligence (AI) within electronic commerce platforms has drastically changed the landscape of the consumer market in India, as the algorithmic systems now dictate product visibility, pricing frameworks, recommendations and targeted advertising which greatly informs consumers’ buying behaviour. Although these innovations provide consumers with increased efficiency, ease of use, and personalisation, they also encourage the manipulation of consumer behaviour in covert, yet very effective ways, while introducing a myriad of complex questions surrounding the accountability and liability for products created by AI technology.

In addition, the Indian consumer protection framework, governed mainly by the Consumer Protection Act of 2019 and the Consumer Protection (E-Commerce) Rules of 2020) was developed with the assumption that consumers are human beings who have agency and can make informed choices, and are always able to view and assess a product before purchasing it. This article examines whether the current Indian legal framework can effectively address consumer protection concerns arising from the use of AI in electronic commerce. In particular, the author will identify significant regulatory “blind spots” regarding the lack of transparency of algorithms, attribution of liability, and the capacity for enforcement by regulatory bodies and governmental institutions.

Ultimately, the author will use doctrinal analysis of Indian case law, consumer forum cases, and the comparison with international regulatory models, to propose changes to the current Indian framework to develop an effective consumer protection framework that reflects the realities of technology in the global market and adequately protects consumers autonomy within this evolving digital marketplace. Keywords: Artificial Intelligence; Digital Market Regulation; Algorithmic Governance; Consumer Protection Law; Platform Liability

I. INTRODUCTION

E-Commerce is an extremely critical part of the Indian digital economy and has changed how buyers and sellers interact with one another marketing, distribution, and consumption of products and services. With the advent of online store sites such as Amazon and Flipkart, ecommerce acts as a conduit for consumers to connect with multiple sellers in different areas of the globe. However, these companies aren't just neutral platforms facilitating transactions; they are now actively organizing and shaping consumer preference through the use of AI and algorithms that dictate how products are displayed and ranked, what price points are offered, and which advertisements get shown. 1 The emergence of AI and algorithm-based decision systems represents a disturbing shift in the balance of power between sellers and buyers.

Traditionally, consumers exercised their power of choice through physical comparison, negotiation, and firsthand inspection of products. On the other hand, buyers using online platforms are subjected to a predetermined field of choices before the actual purchase takes place. The algorithm selects what products a consumer will see first, which products are perceived as affordable, and which products are thought of as popular or trustworthy. Thus, consumers are increasingly making their decisions based on the artificially constructed or designed economy of the online retailer instead of exercising an independent evaluation of all possible options on their own merits. 2 The Indian consumer protection laws have traditionally focused on consumer welfare through correcting information asymmetries and power imbalances in negotiations.

The Supreme Court has consistently stated that consumer legislation should be interpreted under its intended purpose, which is to enhance consumer welfare and to hold powerful market participants accountable for their actions. Artificial intelligence, deployed through e-commerce, creates a new asymmetry that is different from the traditional ones in that it relies on predictive behavioral influence and automated decision-making instead of traditional informational imbalances. The Consumer Protection Act was passed in 2019 to bring Indian consumer law up to date with modern standards of enforcement, as well as to create a statutory regime of product liability for consumers. Additionally, the Consumer Protection (E-Commerce) Rules were enacted in 2020 1 Standing Committee on Consumer Affairs, E-Commerce-Issues and Challenges Lok Sabha Secretariat, 2022

2 Consumer Protection Act, 2019, No. 35, Acts of Parliament, 2019 (India)

to recognise the unique nature of digital marketplaces and the risks they present to consumers.3 However, even though the laws have advanced in some areas, they continue to be reactive and largely unsuccessful at addressing issues associated with algorithmic consumer harm. This article, therefore, examines whether or not the consumer protection framework in India is sufficient when examining issues related to AI-driven e-commerce, including the manipulation of consumers and the absence of product liability when products are sold through automated systems.4

II. ALGORITHMIC GOVERNANCE IN E-COMMERCE MARKETS

E-Commerce platforms depend heavily on machine learning systems that analyse large amounts of consumer data, for instance, browsing behaviour, purchase behaviour, demographic data, etc. , to make predictions about consumer preferences and optimise desired outcomes from a platform, including engagement, conversion rate, and profitability. By using "feedback loops" to continuously improve their performance and provide a more accurate representation of consumer preferences, algorithms dynamically adapt to changes in the market and use new data to improve their performance over time. The use of algorithms represents a fundamental shift in the way that markets operate. Decisions that were previously made by human beings, such as what products to feature/promote and how to price them are now being made by machines.

These algorithms operate at massive scale and speed, and frequently lack clearly established guidelines. Users are rarely provided with any insight regarding the reasoning behind algorithmic recommendations or pricing, which leads to a lack of transparency and creates significant information asymmetries between the platforms and the users. 5 The lack of clarity behind how algorithmic decision-making occurs prevents consumers from being able to understand if they are treated fairly, discriminated against, or exploited. While an unfair trading practice is clear and can be detected easily, the influence of algorithms exists inherently within the platform design and can be difficult to identify and to regulate. Indian consumer law allows room for addressing these issues. In Indian Medical Association v V P 3 Consumer Protection (E-Commerce) Rules, 2020, Gazette of India, Extraordinary, pt.

II, sec. 3, sub-sec. (i)

4 Central Consumer Protection Authority, Guidelines for Prevention of Dark Patterns (2023)

5 Hal R. Varian, Artificial Intelligence, Economics, and Industrial Organization, Nat’l Bureau of Econ. Rsch. Working Paper No. 24839 (2019)

Shantha,6 the Supreme Court used a broad definition of the word service to include all types of professional and technical services to be covered under the consumer protection laws. Therefore, the algorithms that affect consumers’ purchasing decisions through recommending products or adjusting prices on their behalf would also be subject to the same protections as other commercial services offered by an individual or a business in India. A key difference in analysing algorithmic versus traditional methods of unfair trade practices

is the way in which a series of behaviours and transactions reveals unfairness to regulators. The behaviours and transactions resulting in traditional unfair practices are episodic and verifiable externally, thereby allowing regulators to impose punishment immediately after the unfair behaviour or transaction has occurred. Algorithmic systems create systemic relationships between consumer behaviours and the actions of the platform owner; therefore, the platforms are the driving force behind the consumer's behaviour, thus impacting the way and nature of the way consumers are influenced. The influence caused by algorithmic systems cannot be attributed to a single specific wrongful act, as consumers will continue to be exposed to continuous algorithmic influence through the platform, making it impossible for regulators to pinpoint the source of consumer harm without due diligence.

There is now a regulatory blind spot in how consumer harm occurs, as there are no clear evidentiary triggers to support the existence of consumer harm. Hence, the assessment of both fairness and accountability under these circumstances must be redefined. 7 III.

CONSUMER

MANIPULATION

THROUGH

ALGORITHMIC

PERSONALISATION

The way consumers are manipulated in marketplaces that utilise artificial intelligence represents a fundamental change from the traditional approach to consumer protection law, as it is not based on deception or misrepresentation, as is the focus of traditional regulatory systems. The traditional regulatory systems were designed to address and provide a remedy for the harm to consumers from false or deceptive marketing, whether that be through overt misrepresentations in advertising, such as misleading statements, deceptive marketing practices, such as misleading advertisements, or through material omissions that negatively impacted the consumer's ability to make an informed choice. AI-mediated markets, in contrast,

6 Indian Med. Ass’n v. V.P. Shantha, (1995) 6 SCC 651

7 Id

are oriented toward facilitating consumer behaviour and actions by using behavioural nudges, covert influences rather than blatant misstatements or misrepresentations. Because platforms utilise the structure of digital choice environments to facilitate consumer behaviour without using any overt predictions or direct coercive tactics, there is no clear distinction between deceptive practices defined by overt misrepresentations and influence practices defined by manipulating behaviour when addressing the issue of consumer harm. In addition to complicating regulatory responses, the implementation of behavioural economics within the design and implementation of algorithms further complicates the ability of consumers to have control over their own choices and decision-making processes.

Increasingly, e-commerce platforms are using predictive analytics to take advantage of biases that could potentially lead to negative consumer experiences. 8 Many cognitive biases, such as loss aversion, scarcity effects, and impulsivity, are now a part of E-Commerce platform designs. This means that biases will no longer occur naturally in the marketplace but will instead be intentionally utilised to maximise platform engagement and conversion rates. Therefore, rather than allowing consumers to make informed purchases based on logical thinking, consumer decisions are increasingly being influenced and coerced by behavioural nudges rather than logical thinking, which has significant implications for reducing consumer autonomy within the digital marketplace.

9 Using personalisation, platforms can engage in real-time targeted marketing at a level never before possible and, as a result, amplify the imbalance between consumer and platform. Dynamic pricing, recommendations, and promotional messages can be tailored according to predictions on how the consumer will behave and delivered when the consumer is likely to be at their most vulnerable. As with many practices associated with the increased convenience and relevance of marketing, consumers do not realise the extent of the data being used to influence their purchasing environment.

The lack of transparency associated with the practice does not facilitate informed consent; as such, this failure to disclose undermines the consumer's ability to make an informed choice and violates the core consumer protection principle that choices should be made freely and with an understanding of the impact and implications of doing so. 10 Indian consumer jurisprudence provides a legal basis for the recognition of algorithmic

8 Ryan Calo, Digital Market Manipulation, 82 Geo. Wash. L. Rev. 995 (2014)

9 Id

10 Organisation for Economic Co-operation and Development, Artificial Intelligence and Consumer Policy (2020)

manipulation as a legitimate harm to consumers. In the case of Lucknow Development Authority v M. K. Gupta,11 The Supreme Court of India held that the injury to a consumer extends beyond a defect in a physical product; that consumers may suffer from mental harassment and economic loss because of unfair practices. By interpreting the consumer law in this way, it provides an avenue for the legal protection of a consumer from the intangible and psychological harm associated with digital manipulation, and therefore, the jurisprudence supports the classification of algorithmic manipulation as a form of unfair trade practice. The main concern for regulators is determining who will be held responsible for any damages caused by autonomous systems.

Algorithms are typically deployed without any human involvement, meaning they will continue to grow and evolve through their ongoing relationship with each new user. Because an algorithm's intent is not always obvious, it becomes difficult to apply the current consumer protection laws which look to fault-based standards. 12 If companies are able to utilize algorithms to avoid being liable for damages, this will ultimately defeat the underlying objectives of consumer protection laws. Therefore, the focus must be shifted from evaluating the intent of an algorithmic practice towards evaluating the effects of these practices on consumer decision-making. 13 Additionally, the ongoing use of algorithms to manipulate consumer behaviour remains intact and reinforces the existing structural inequities that exist within the digital marketplace.

Examples such as personalised pricing and targeted recommendations could provide consumers with different services based on their perceived characteristics, such as spending capability, geographical location, or propensity to buy certain items. With the lack of transparency and oversight, consumers are incapable of contesting the disparate treatment they receive from businesses. By recognising algorithms as a means of perpetrating a systemic unfairness on consumers, we have an obligation to protect consumers as individuals, but also to maintain fairness in the digital marketplace and support public confidence in the digital economy. 14

11 Lucknow Development Authority. v. M.K. Gupta, (1994) 1 S.C.C. 243

12 Cary Coglianese & David Lehr, Transparency and Algorithmic Governance, 71 Admin. L. Rev. 1 (2019

13 Christoph Busch, Liability for Artificial Intelligence Systems under EU and Comparative Law, 58 Common Mkt. L. Rev. 87 (2021) 14 Shoshana Zuboff, The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power (PublicAffairs 2019)

IV. PRODUCT LIABILITY UNDER CONSUMER PROTECTION ACT, 2019

The introduction of the Consumer Protection Act, 2019, has changed the landscape of consumer law in India by creating a unified statutory scheme in relation to product liability. Before the Act, the approach taken by Indian courts had been fragmented in terms of remedies available to consumers against products they purchased. Therefore, the Act’s provision extends the application of liability beyond just the manufacturer to also include not only sellers of products and service providers but also others involved in the distribution of products. With the expansion of liability towards more than just one individual, this represents both the growing nature of today’s marketplace as well as how it works in digital/e-commerce environments where consumer injuries may be the result of actions taken by multiple parties, including platforms, intermediaries, and technology.

The legislature’s intention for expanding the number of possible liable actors was to ensure that there would be available remedies for consumers who suffered injury and to ensure that those individuals were not able to escape liability because of the complexity that arises with commercial relationships. 15 The difference in structure between the traditional liability concept and algorithmic systems causes serious tension within doctrinal areas of law. Consumers will have a difficult time establishing causation or identifying a plaintiff for AI-related damages because consumers must rely on multiple layers of processing data and make automated inferences. Therefore, using traditional standards of evidence may not be sufficient to identify the source of harm caused by AI.

Although the Consumer Protection Act clearly defines “E-Commerce” platforms as product sellers, it does not provide any specific direction regarding the assignment of liability for AI-decision making that occurs on those platforms. Therefore, there remains uncertainty regarding whether E-Commerce platforms can be held accountable for the algorithmic outcomes that materially affect the consumer’s well-being. 16 As a result of this uncertainty, consumers are typically considered the victims of algorithmic governance; they suffer the damages associated with the ambiguous and constantly changing systems of algorithmic governance while having limited legal avenues available to seek compensation. This creates a result that is inconsistent with the policy interests of product liability law that includes placing the risk of harm on those best able to prevent such harms and

15 Consumer Protection Act, 2019, No. 35, S.2(34), 2(37), 83–87, Acts of Parliament, 2019.

16 Christoph Busch, Liability for Artificial Intelligence Systems under EU and Comparative Law, 58 Common Mkt. L. Rev. 87 (2021)

absorb the costs associated with them. Without clear legal guidance on algorithmic liability, the promise of consumer protection found in the Consumer Protection Act, 2019, is only partially fulfilled. Therefore, addressing the lack of legal clarity regarding algorithmic liability is critical to the goals of this act.17

V. PLATFORM ACCOUNTABILITY AND INDIAN CASE LAWS

The way Indian Courts have viewed E-Commerce platforms shows that the courts are slowly starting to move away from the way they used to determine how companies are classified as intermediaries and are moving towards treating E-Commerce platforms as active participants in their respective markets. The case of Amazon Seller Services Pvt Ltd v Amway India Enterprises Pt. Ltd,18 decided by the Delhi High Court demonstrated this change in approach because it acknowledged that online E-Commerce platforms, such as Amazon, exercise significant control over market outcomes since it has control over the products consumers can see, who can sell their products through E-Commerce websites, and the conditions by which individuals can participate in the E-Commerce marketplace in the first place.

Additionally, the reasoning in this case demonstrates a move away from the contract-based view of platform participation and instead looks at the functional impact an E-Commerce platform has on shaping consumer decisions and the overall makeup of an E-Commerce marketplace. In a similar case, Flipkart Internet Pvt Ltd v State of Tamil Nadu,19 The Madras High Court similarly found that E-Commerce platforms control pricing for sellers through both technological and operational means, thus indicating that E-Commerce platforms should not have a presumption of neutrality, and therefore, digital marketplaces cannot avoid regulatory scrutiny simply by relying on their status as intermediaries to escape regulatory oversight.

Collectively, analyses in these two cases indicate that there is a growing consensus among the judiciary that E-Commerce platforms need to be evaluated based upon the economic substance of their market power and not based solely upon the labels attached to their business models. Consumer fora continue to solidify this changing view from the judiciary by assigning liability on the platforms that facilitated their operational involvement, and so resulted in consumer harm. In Shaik Umar Farooq v. Flipkart Internet Pvt. Ltd. ,20 the District Consumer Disputes

17 Christopher Hodges, Product Liability: Problems and Reform (Oxford Univ. Press 2019)

18 2020 SCC Online Del 454

19 2021 SCC Online Mad 3325

20 2022 SCC Online NCDRC 519

Redressal Commission found the platform to be jointly liable, along with the seller, as a result of the seller overcharging the consumer, and specifically rejected the defence of an intermediary supplier. This case is of particular relevance in that it was held that systems designed by the platform, which include pricing interfaces, as well as transactional controls, can directly enable unfair business practices. The reasoning in this case is in line with the consumer protection objectives of consumer law, which are intended to prevent larger and more powerful actors in the market from shielding themselves from liability due to complex contractual and technological relationships that insulate them from the consumer’s protections.

However, Indian Courts have not yet been directly presented with the question of whether algorithmic decision-making creates a basis for determining liability or not. Currently, Indian Courts appear to focus on the control of a structure or of the market, whether in respect of access to sellers, influence over price, or contractual dominance, and not to focus upon the autonomous nature of the algorithmic systems in the way that they produce recommendations, rankings and or tailored offers for consumers or in the context of the prevailing market going forward. Therefore, the current judicial thought on this subject has not developed to the extent that it addresses the legal questions of causal effect, attribution of liability and of responsibility for conduct that is the result of a non-transparent, self-learning algorithm rather than the conduct of a human being.

Regulating AI-based E-Commerce effectively reveals a disparity between the complexity of algorithmic systems and the institutional capability of consumer protection agencies in India. Historically, consumer forums were designed to resolve disputes concerning physical goods sold by identifiable sellers with obvious unfair business practices involved. Algorithmic systems, however, do not operate in such a straightforward manner; they instead rely upon an algorithmic methodology that is opaque and based on data sets, which necessitate the access of proprietary data sets, knowledge of technical areas, and an understanding of multiple disciplines including computer science, economics, and law, to understand how algorithmic decision-making has caused harm to consumers.

Furthermore, consumer protection adjudicatory bodies have a limited ability to correlate consumer harm with its source, as the cause of damage from automated decision-making is often distributed across several layers of automated inference, rather than a single action taken by a seller. 21

21 World Econ. Forum, Global AI Governance Framework (2021)

VI. COMPARATIVE JURISDICTION

Jurisdictions in other locations have attempted to adapt more proactively and predictively through regulatory measures, especially as it pertains to how AI technology affects consumer behaviour and relationships, with European Union member states being the most prominent of those jurisdictions. Through its Digital Services Act (DSA), the European Union has implemented a transition from reactive regulation to pre-emptive or “ex ante” regulation of digital platforms. The DSA has determined that recommendation systems are significant for influencing consumer decisions and, as such, set forth defined and specific guidelines with which the largest online platforms must comply in terms of transparency obligation regarding the key factors on which recommendation algorithms are based.

Further, the DSA requires that meaningful modifications or opt-out settings be provided to individuals who receive personalised recommendations from these online platforms. The DSA intends to fix the imbalance of information between online platforms and consumers by addressing the design and functionality of the interface prior to the point of established consumer detriment or disadvantage through algorithmically driven content or product recommendations; thus, restoring a measure of consumer control over the algorithmic influence. 22 The DSA also prohibits manipulative interface designs, which can impair or distort consumers’ ability to make decisions, not just by having manipulative interfaces that trick consumers into doing something that they would not normally do.

The DSA’s prohibition of manipulative interface designs is an example of a regulatory approach based on recognising that consumer harm is more than just isolated deceptive acts, but rather on systemic and pervasive design choices built into the architectural design of the platform. 23 By treating manipulative interface design as a structural issue and not a single instance of transaction disturbance, the DSA provides consumer protection through the regulation of markets, where EU member states will have market governance authority rather than the governments of consumers and regulators, who would have difficulty detecting algorithmically driven unfairness on an ex-post basis.

24 In contrast to the EU, which has a comprehensive ex-ante regulatory framework for AI, the US has adopted a more enforcement-based approach, focusing on utilising existing consumer 22 Regulation (EU) 2022/2065, of the European Parliament and of the Council of 19 October 2022 on a Single Market for Digital Services (Digital Services Act), 2022 O. J. (L 277) 1 23 European Commission, Guidance on the Application of EU Consumer Law to Online Interface Design (2021)

24 Org. for Econ. Co-operation & Dev., Dark Commercial Patterns (2022)

protection laws to regulate AI-related activities. The FTC has been clear in its statements on this matter that AI itself does not create a regulatory void, nor does it lessen the legal requirements under the FTC Act that prohibit unfair and deceptive acts or practices. The FTC has issued guidance for businesses and taken enforcement actions that indicate companies using artificial intelligence must ensure their AI systems do not result in misleading consumers, discriminatory outcomes, or other forms of unfair harm. The FTC’s position demonstrates an institutional commitment to technological neutrality; therefore, legal responsibilities apply to harmful behaviour regardless of whether the harm originates with an individual or an automated system.25 A regulatory approach based on enforcement reflects a pragmatic regulatory response to

technology. In future use of technology, the U. S. regulators are planning to apply existing consumer protection laws to emerging technologies and their associated harms, rather than creating an entirely new regulatory framework for Artificial Intelligence (AI) as currently exists in the U. S. under MLA (Machine Learning Act) 1999. However, using terms like “unfair”, “deceptive”, or other terms within existing laws does provide for a broader range of algorithmic-harm. As technology evolves rapidly, it is very important to be flexible in regulatory action, therefore creating a prescriptive regulation could potentially create a situation where the regulation would become irrelevant due to the pace of technological advancement. At the same time, the U. S.

tends to rely on ex-post enforcement, meaning current regulations are generally enforced after the harm to consumers has taken place as opposed to, or in conjunction with, operating under a pre-emptive policy regime. 26 When combined, the two models indicate contrasting strategies for dealing with algorithmically-induced consumer harm. The ex-ante means protects against future systemic risk by creating certainty through regulation, whereas the ex-post approach gives the Police more freedom to respond to new forms of wrongdoing and provides a level of flexibility to the Police. A blended regulatory model that combines both elements of these models would provide the best means of protecting consumers in markets that are AI amplified.

This approach does not treat algorithmic influences as an isolated technical matter, rather it considers the impact of algorithmically driven harm as a significant aspect of modern consumer law. 27

25 Fed. Trade Comm’n, Protecting Consumers in the AI Economy (2021).

26 Cary Coglianese & David Lehr, Transparency and Algorithmic Governance, 71 Admin. L. Rev. 1 (2019).

27 OECD, Artificial Intelligence and Consumer Policy (2020).

In particular, the lessons learned by the EU and the US can help inform India's approach. Currently, India's consumer protection framework is primarily complaint-driven and reactive, with little legal recognition of the existence of algorithmically driven systems. Evidence from these two jurisdictions indicates that to adequately protect consumers in the digital economy it is necessary for legislation and regulation to offer consumers both proactive protections and effective enforcement. To improve India's framework and ensure adequate consumer protections are provided, India will need to establish specific transparency obligations for recommending systems, delineate the accountability of platforms for algorithmically driven outcomes, and empower regulators to ensure regulators have both the authority and expertise to respond to harms caused by AI.28

VII. RECOMMENDATIONS

India needs to go beyond reacting to consumer protection issues and create a new regulatory framework that properly reflects the impact of AI in the marketplace. The current consumer protection laws in India assume that harm resulting from consumer products is due to human actions or specific defects in physical products. This is no longer accurate in AI-driven or digitally enabled markets where the automated systems are creating consumer experiences or pricing options, or influencing consumer choices. Therefore, the law needs to provide statutory recognition of the potential for harm caused by algorithmic services. Providing this kind of recognition will ease the burden of proof that consumers have to bear when attempting to prove that an opaque algorithmic process caused him or her harm, something that is extremely difficult given the complexities involved.

By expressly incorporating algorithms as part of the scope of consumer protection laws, legislatures can make sure that future technological advancements do not permit businesses to avoid responsibility for their actions because of the inherent complexity of technology. The transition from optional to mandated transparency over the workings of recommender systems and dynamic pricing algorithms is part of creating consumer rights. While consumers can't see the underlying logic of the systems that recommend and provide pricing information, consumers are entitled to be informed about the most important factors that go into determining what they see, as well as meaningful options to request changes to those suggested products and services or to refuse personalisation altogether. Viewing transparency as a consumer right

28 NITI Aayog, National Strategy for Artificial Intelligence: #AIForAll (2018).

rather than an optional, benefit reinforces the underlying tenet that the foundation of consumer protection is informed decision-making. It is also important to define the extent to which platforms can be held liable for any harm produced as a result of automated systems that they design, build or benefit from. In such instances, platforms should not be allowed to escape liability by accusing themselves of only acting as facilitators. Certain laws should clearly outline what level of accountability a platform has regarding its actions so that there is no ambiguity or confusion, preventing regulatory arbitrage, and placing liability on parties able to adopt and manage risk and prevent such harm. In addition to these legal changes, we must also strengthen institutions.

Consumer protection agencies will be ineffective in resolving disputes involving the use of AI or algorithms if they do not have the appropriate access to technical knowledge. Overall, empowering and improving the capabilities of consumer agencies, through appointing technical advisers and providing training to adjudicators and access to dedicated investigative support, must occur in order to allow consumer agencies to analyze algorithmic evidence appropriately and promptly. Absent such reform, even the best legal standards will remain unenforced. These measures will enable India's consumer protection system to keep pace with technological development and protect consumers' interests and promote fairness in the digital economy.

VIII. CONCLUSION

Artificial Intelligence Models have changed the e-commerce infrastructure and thereby changed how consumers interact within the consumer markets of India. Today, the digital platforms that facilitate E-Commerce transactions for users are not just neutral third parties; rather, they are now active and powerful intermediaries shaping user decisions through an algorithmic system that determines what products have the highest probability of succeeding. There are many benefits that these systems provide; however, they also create new ways to exploit consumers and provide a new way to take advantage of users through methods that the existing laws were not designed to cover.

This article illustrates that even though India’s Consumer Protection Act of 2019 improves and strengthens consumer remedies, it does not incorporate the concepts of algorithmic governance that have developed; therefore, the existing legal framework still has many gaps regarding transparency, attribution of liability, and institutional capacity. Courts in India have acknowledged the significant amount of power and control e-commerce platforms have over

consumers and have begun to develop new approaches to addressing this power instead of following a strict intermediary model; on the other hand, the courts still do not have enough involvement with the companies’ algorithmic decision-making processes, leaving many consumers unprotected from algorithmically determined harms.