New Delhi. As India enters the festive shopping season, the growing use of artificial intelligence (AI) agents in online shopping could bring a major change to the way consumers make purchases. AI is no longer being used only to identify products or compare prices. New AI agents are being developed that can understand a customer’s requirements, search for products and sellers, compare prices and potentially complete purchases on the customer’s behalf. Banks and financial institutions have raised concerns over fraud, data privacy, payment authorisation and responsibility when an AI system makes a mistake.
Online shopping typically rises around Dussehra and Diwali in India. During major e-commerce sales, customers compare prices offered by different sellers, bank discounts, delivery timelines and product reviews. If AI agents begin handling a significant part of this process themselves, payment systems will face questions over what exactly the customer authorised and which decisions were independently made by the AI agent.
Proposal for Conducting Cyber Crisis Drill, Tabletop Exercise (TTEx) & CCMP Readiness Exercise
AI moves from recommendations to purchases
AI in online shopping was initially used mainly for providing product information, comparing prices and making recommendations. Technology companies are now developing agents capable of handling several steps on their own based on customer instructions. A customer could specify their requirements, budget and preferences, after which an agent could search available products and compare different options.
The next stage could see the same agent select a product and seller and initiate or complete the purchase process. This could reduce the need for customers to click through every stage themselves. However, the increased convenience also raises an important question: if an AI selects the wrong product, chooses an unsuitable seller or spends more than the customer intended, who will be responsible?
In traditional online shopping, the customer selects a product, enters payment information and authorises the transaction. With an AI agent, the customer may provide only an initial instruction while the software handles several subsequent steps. This could make it more complicated to establish the customer’s actual consent when a payment is disputed.
Fraud risks involving AI already exist
Indian banks have already been warning customers about cyber fraud enabled by AI. Criminals are using voice cloning, realistic-looking videos, fake conversations and personalised messages to gain victims’ trust. AI-generated content can make fraudulent communications appear more convincing.
During the festive season, fake discounts, gifts, prizes and shopping offers are also commonly used by cybercriminals. Messages on WhatsApp, Telegram, Instagram and other platforms can direct users to suspicious websites or malicious links.
Customers are advised to check website domains, avoid unknown links and never share card numbers, CVV, PINs, net-banking passwords or transaction OTPs. However, AI agents could create a different kind of risk because a customer may initiate a transaction through an AI system without personally visiting a suspicious website.
Consumer trust remains a key issue
A global consumer study covering 28,000 people across 17 countries found that 82 per cent of respondents had used AI. At the same time, 50 per cent said they would prefer to tell an AI what they need, receive options and complete a purchase in one place. However, 58 per cent said they cross-check AI recommendations before making a purchase.
The figures indicate that consumers may be relatively comfortable using AI as a shopping assistant, but handing over complete control of the purchasing process remains a more cautious step. AI-based automated shopping is expected to emerge initially in frequently purchased and relatively lower-risk categories, where consumers may be more comfortable delegating decisions.
Payment and responsibility questions for banks
For banks, the challenge is not simply whether an AI agent can make a payment. The bigger question is whether such payments can be authenticated, monitored and disputed in the same way as transactions initiated directly by customers.
Globally, some banks have called for consumers to be clearly informed when an AI agent is involved in a transaction. Greater transparency around how AI makes decisions and stronger protection of customers’ financial data are also being sought.
If an agent purchases the wrong product or takes an unauthorised action, another question arises over how responsibility should be divided among the bank, customer, AI service provider, e-commerce platform and seller. Payment systems may need new security standards and monitoring mechanisms to handle such transactions.
Festive season could test AI shopping
This year’s festive season could become a major test for AI-based shopping. Consumers may increasingly use AI to find deals, compare prices and select products. At the same time, e-commerce companies and payment institutions will need to understand how AI agents interact with their platforms and what permissions they have.
Banks will have to distinguish between legitimate AI-assisted transactions and potentially fraudulent activity. E-commerce companies will need mechanisms to identify authorised AI agents and determine the limits within which they can operate. For consumers, the key question will be how much control they retain once AI moves beyond making recommendations and starts taking actions on their behalf.
AI can influence shopping even before payment
The study found that 58 per cent of consumers check AI recommendations before purchasing, while 46 per cent said AI recommendations had changed the brands they considered. This indicates that AI’s influence may begin well before the payment stage. It can shape which products, brands and sellers make it onto a consumer’s final shortlist.
Therefore, AI-based shopping raises questions not only about payment security but also about the transparency and reliability of recommendations. Consumers may need to understand why a particular product was recommended and how much control they retain over the final decision.
As online shopping increases during the festive season, the use of AI agents could also expand. While the technology may make shopping easier and faster, banks and financial institutions will have to develop new approaches to fraud prevention, customer authorisation, data protection and dispute resolution as AI evolves from a shopping adviser into an agent capable of taking action.
About the author — Suvedita Nath is a science student with a growing interest in cybercrime and digital safety. She writes on online activity, cyber threats, and technology-driven risks. Her work focuses on clarity, accuracy, and public awareness.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
