Introduction
Commerce is moving towards hyper-personalization and an experience-driven model where interaction, search, and purchase blend into one process. The conventional separation between chatting, browsing, and shopping is fading as artificial intelligence becomes central to online retail.
India has now entered this new phase of digital transformation through Agentic Payments and Instant Checkouts. The National Payments Corporation of India (NPCI) and fintech firm Razorpay have partnered with OpenAI to pilot AI-powered payments on ChatGPT. The project includes Axis Bank and Airtel Payments Bank as banking partners, while Bigbasket, owned by the Tata Group, will be among the first platforms allowing customers to shop directly through ChatGPT.
This innovation enables users to search for products, compare options, place orders, and make payments without leaving the chat interface. It marks a major step toward “conversational commerce”, anchored in OpenAI’s Agentic Commerce Protocol (ACP), which integrates discovery, decision-making, and payment into a single interactive experience.
Although the Indian pilot is at an early stage, ChatGPT’s Instant Checkout feature is already functional in the United States. Through collaborations with Stripe and Etsy, users can browse, select, and complete purchases within the chat. The current model supports single-item purchases from platforms such as Etsy and Shopify. OpenAI plans to extend this to multi-item carts and link with nearly one million Shopify merchants.
This expansion signals a structural shift in e-commerce where conversations replace traditional interfaces. Instead of navigating multiple pages, consumers can now complete their entire purchase journey in one dialogue. The chat itself morphs into a marketplace.
How Does It Work?
In the US version of Instant Checkouts, the process unfolds as a natural conversation between user and AI.
A shopper might begin with a request such as “best running shoes under $100” or “gifts for a ceramics lover.” ChatGPT responds with relevant product options within the chat. Users can compare specifications, refine choices by price or preference, and interact further to identify the best match. The AI acts as a personal shopping assistant that understands context and adjusts recommendations accordingly.
Once a product is selected, the transition to payment is seamless. Users simply click the “buy” button and confirm their shipping and payment details. ChatGPT then processes the order and completes the transaction using available methods such as credit and debit cards, Apple Pay, or Google Pay.
Through this entire process, the customer never has to leave the chat. The AI agent does the entire process seamlessly while giving an interactive and personalized experience.
AI-Driven Future of Online Shopping
Traditional online shopping involves multiple steps: searching for a product, browsing through links, visiting different websites, and navigating separate interfaces for checkout. Even when platforms added chatbots to assist in comparison, payments remained external.
Agentic Commerce changes this structure. Through OpenAI’s ACP, it creates a unified and conversational shopping experience that recalls user preferences, past purchases, and contextual information. It replaces fragmented browsing with a single, continuous interaction that sustains user engagement.
Competitors such as Google and Perplexity AI have also begun experimenting with AI-based payment services. Perplexity’s system is currently available only to Pro users in the United States, while Google’s version is still in development. This suggests that conversational commerce is not a passing trend but a lasting shift in how consumers will engage with digital markets.
Implications for Businesses and Consumers
OpenAI estimates that 700 million users engage with ChatGPT each month. With agentic payments, these everyday users could also make purchases during routine conversations.
For businesses, the model presents both opportunity and disruption. The ease of conversational transactions could increase sales, reduce cart abandonment, and bring smaller merchants closer to potential customers. At the same time, success will depend on how effectively sellers adapt their product data for AI-based systems. Businesses will need to ensure that their products are visible and optimized for AI discovery. The ability of algorithms to interpret catalogues and identify relevant listings will determine how easily consumers encounter a brand.
Consumers stand to benefit from greater convenience and personalization. Yet this simplicity introduces new concerns about fairness, privacy, and regulation. As the technology scales, its governance will become as critical as its innovation.
Regulatory and Ethical Concerns
1. Transparency and Fair Competition
Questions arise about how AI agents prioritize product listings. Critics worry that AI may favor certain brands or sponsored results without clear disclosure, creating potential competition issues. OpenAI maintains that product rankings are “organic and unsponsored, based purely on relevance to the user.” It further clarifies that Instant Checkout availability does not influence ranking; rather, parameters such as price, availability, quality, and seller status determine the final list.
2. Customer Experience and Accountability
Another concern involves order management, refunds, and delivery tracking. After placing orders through chat, customers may naturally expect to receive all updates and notifications within the same conversation. At the same time, retailers face the challenge of verifying customers and confirming orders solely through a chat interface. Such operational issues require clear protocols and standardized integration between merchants and AI agents.
3. Privacy and Data Protection
Personalization relies heavily on user data. The chatbot accesses contextual information such as prior memory and custom instructions to refine recommendations. However, amid the ease of casual conversation, users may not fully realize the extent of data being shared during these interactions. This raises privacy and consent challenges, necessitating stronger disclosure mechanisms and compliance with data protection laws.
4. Consumer Protection and Accidental Orders
The conversational nature of transactions can blur intent. A casual query could inadvertently trigger a purchase. Traditional e-commerce offers several review checkpoints before payment, while conversational commerce reduces these friction points. This highlights the need to establish effective safeguards against accidental or unauthorized orders.
5. Returns, Modifications, and Dispute Resolution
Despite assurances that the system is built for trust, practical questions and lingering uncertainty about how agentic payments function still remain. How will returns and cancellations function in this model? What if recommended products go out of stock? How will subscription or recurring payments be managed? These scenarios require well-defined consumer protection frameworks suited to conversational interfaces.
6. Cross-Border and Fraud Risks
As digital commerce expands across borders, businesses face complex compliance challenges, while at the same time, advances in technology continue to fuel the evolution of sophisticated fraud techniques. Traditional fraud detection depends on identifying human anomalies, but in agentic commerce, unusual agent behavior may still be legitimate. This necessitates multi-agent behavioral analytics and real-time explainability, capabilities far beyond current rule-based systems.
Conclusion
Retailers and consumers in India had only recently grown comfortable with personalized ads and quick commerce, and agentic commerce accelerates that transition. The next phase of digital retail will not simply depend on smarter algorithms, but on systems of accountability that preserve transparency and trust. Agentic commerce is not only a technological breakthrough; it marks the beginning of a new regulatory frontier for digital markets.
Authored by Viplavi Vinod Joshi, Consultant Economist at The Dialogue