Algorithmic price discrimination: The hidden biasness

July 3, 2026

Author: Bhavna, 2nd-Year Law Student (Undergraduate), Faculty of Law, University of Allahabad, Prayagraj, Uttar Pradesh.

Table of Contents

Algorithm

An algorithm is a set of rules or instructions given to an AI, machine learning program or automated system to help it learn it on its own – and solve problems.

In a simpler sense we can say that the input provided to the system is processed and stored along with the outputs which results from following specific procedure, then these stored data are used to produce further results with similar inputs. This whole procedure is stored and used to train further models thus automatizing the process.

Algorithms help us to personalize our experience but have you ever wondered about the problems caused by it? Let’s dive into one of the critical and significant aspects of algorithms.

Algorithmic discrimination

When a certain group is at disadvantage due to the algorithm systematically producing outputs based on certain characteristics i.e.gender, race, etc. which might put them into certain disadvantage, we call it algorithmic discrimination. It works on the data that are fed into systems thus producing outputs based on it.

Direct discrimination algorithm- When protected attributes are used as an input variable, it is a direct discrimination algorithm. E.g. gender in insurance pricing.
Indirect discrimination algorithm- When proxy variables such as zip code, education etc. correlate with protected attributes thus producing discriminatory outcomes without explicit use.

Introduction

Algorithmic price discrimination refers to the changing of prices based on profiling the user thus characterising the customer. Unlike dynamic pricing which depends on supply and demand, it functions on the customers willingness to pay. It showcases the varying prices to the varied customers based on vast data available about them using an algorithm with the aim of charging the customers the maximum price they are willing to pay for goods or services. It jeopardizes ethics of privacy, equality and trade. The concept of algorithmic price discrimination could be best explained through following illustrations:

Illustration 1: There is a difference in pricing of flights based on frequency of travelling, income, time of booking etc.
Illustration 2: There is a difference in insurance policies based on age, gender, earning potential etc.

Profiliation of customers

Profiliation of customers is done by tracking their browsing history using cookies, social media activities, or tracing the device type i.e. android or iOS, or the browser they are using such as safari or chrome. These factors are taken as indicators of a person’s purchasing capacity. Algorithms use the data along with human psychology to determine or predict a person’s willingness to buy a product.

Based on this, consumers are segregated into different groups and are shown personalised prices.Using the above mentioned factors, algorithm sets up the price just below the maximum price a person is willing to pay for the product but ridiculously higher than the actual price. Most of the time, the customers incur loss due to this wide margin between the prices.

The problem of dark patterns

Dark patterns can be interpreted as deceptive user interface or user experience used by websites which trick buyers into making choices which they might not have made under usual circumstances eventually profiting the business. Some of the examples of dark pattern are:

False urgency- Creating a scenario of limitation of stock or setting countdown which causes an environment of urgency in the mind of buyers leading to impulsive purchase.
Basket sneaking- It refers to the addition of items or services sneakily into customers’ baskets. Many of the shopping apps and food delivery apps practise basket sneaking. If the customer is not vigilant then the person would face losses.
Tracking walls- It makes data collection and tracking mandatory for accessing certain services. It compromises with a person’s privacy and autonomy.

Algorithmic price discrimination uses such dark patterns to create situations where the customer has only superficial autonomy over his choices and influences customers to buy products at exorbitant prices. Personalised pricing is usually offered as a non-negotiable part of the contract.

The problem of algorithmic collusion

Algorithmic collusion uses different AI models or software to set high prices in businesses, functioning like hub-and-spoke models bypassing human agreement. Such trained AI models cause artificial inflation affecting the wallets of consumers and are difficult to regulate.

It causes price parity as different customers pay different prices for same goods and services under circumstances such as surgency in demand at different hours of the day or popularization of certain products etc.

Problems with algorithmic price discrimination

It throws certain issues in our way. Few of the issues has been discussed here:

Infringement of right to privacy- Article 21 of Constitution of India states that “No person shall be deprived of his life or personal liberty except according to procedure established by law.” In K.S. Puttaswamy vs. Union of India (2017), right to privacy was included in the ambit of article 21 gaining the status of fundamental rights. Algorithmic price discrimination uses personal data and digital footprints which is a direct violation of the right to privacy. Users also lose control over their personal information which is used to influence their economic decisions.

Price parity- Customers face pay parity. They pay different prices for the same kind of services.

In Federation of Hotel & restaurant associations of India vs. MakeMy Trip pvt. Ltd. & ors. (2022), the competition commission of india (CCI) sanctioned MakeMy trip and Goibibo for algorithms to enforce price parity, effectively penalising hotels that offered lowered prices on their own websites or other platforms.

Sociological consequences- It degrades the value of life and choices of an individual to mere set of data for the profiting purpose of businesses. It is invasive in nature and influences our core decisions knowingly or unknowingly reducing our autonomy on our own life.

Legal provisions across Europe

Digital Markets Act (2022)– It was formulated for business giants. It requires the businesses to maintain transparency in methods of pricing and online advertisement. It also restricts businesses from sharing data of the customers across platforms without users specific consent. It also focuses on the parity clause.

Article 101 and 102 of the Treaty on the Functioning of the European Union– It deals with the issue of hub-and-spoke regulated through AI thus taking the issue of algorithmic collusion under its umbrella.

The Eturas case established that the competition laws were applicable to the issue of algorithmic collusion as well.

General Data Protection and Regulation– It not only deals with privacy laws but also regulates automated decision making, reducing the impact of algorithms.

Legal provisions in India

Section 43A of IT act– It deals with compensation for failure to protect data. It holds body corporate liable for negligently handling sensitive data which could have been protected through reasonable security practices and procedures.

Digital Personal Data Protection Act, 2023– It takes into consideration individual privacy rights along with lawful business operations. It includes right to grievance redressal which extends to automated data processing whose rules are still under development. However it doesn’t have explicit laws to deal with profiliation of customers, algorithmic discrimination etc.

Article 14 of constitution of India– The price parity due to algorithmic bias can be challenged under constitutional safeguard of article 14 which deals with equality before law.

Article 15 of constitution of India– It prohibits discrimination. Algorithmic price discrimination can be challenged under the umbrella of this provision.

Conclusion

Algorithmic price discrimination is detrimental to the very aspect of equality, privacy and human dignity. While algorithms help in personalising choices and making human experience enriching, the algorithmic biasness needs to be dealt with care so it doesn’t undermine the virtue of the constitution and humanity. There has been emergence of various apps and sites which compare the prices but they fail to fully solve the problem. There is no law specifically dealing with the issue of algorithmic price discrimination. The legal safeguards regarding the same are inadequate and still under development in India.

Few suggestions that could help in solving this issue are-

Laws for profiliation of customers and algorithmic discrimination- Specific provisions regarding the same needs to be formulated to deal with this issue effectively.
Audit of online platforms- Audits should be conducted to stop misuse of algorithms for price discrimination.
Laws for regulation of information- There is need for stricter laws determining what information can be extracted, stored and circulated about an individual along with limiting its use for profiliation.
Guidelines to reduce tracking- There needs to be proper guidelines for what activity of an individual can be tracked.
Liberty to regulate digital footprints- An individual should be able to know about the history of their digital footprint and have the liberty to regulate it.
Limiting the usage of AI by large corporations- Limitations should be put on use of AI in deals and price aspects of business.
Price capping- The government should make laws related to fluctuating prices online. They should make laws stating a range to which the price of a good can fluctuate that too only on the basis of supply and demand of goods and services.

At last I can conclude by stating that better cyber laws are needed to treat the legal aspect of algorithmic price discrimination. We should ensure that the customers rights are protected but at the same time businesses don’t incur any loss.

References-

Lorem ipsum amet elit morbi dolor tortor. Vivamus eget mollis nostra ullam corper. Pharetra torquent auctor metus felis nibh velit. Natoque tellus semper taciti nostra. Semper pharetra montes habitant congue integer magnis.