InMobi CEO Sounds Alarm – Transparency and Accountability Are Essential for Agentic Commerce

InMobi CEO Emphasizes Importance of Transparency at AI Impact Summit 2026

Agentic commerce must be transparent and accountable, says InMobi CEO

Naveen Tewari, the founder and CEO of InMobi Group says that while agentic AI holds tremendous potential for commerce, companies must ensure these systems operate transparently and ethically

On February 22, 2026, Naveen Tewari, CEO and founder of InMobi Group, spoke at the India AI Impact Summit 2026 in New Delhi about the future vision of agentic commerce and the "transparency" and "accountability" essential for its realization.

Tewari stated, "If you bring AI intelligence to commerce, you can fundamentally change economic growth. But it has to be real," referencing the "distortions" created by digital platforms over the past decade. "Social media has often manipulated attention. Agentic systems must be transparent and accountable," he clearly asserted.

Industry Trends

Agentic commerce refers to a system where AI agents search for products, compare and evaluate them, verify safety, and complete payments on behalf of consumers. According to Mastercard's explanation, simply instructing "buy blue pants under $50" allows AI to handle everything from search to size selection, payment, and shipping address entry — a world that is rapidly becoming reality.

The rapid attention this technology is receiving is driven by successive announcements at India AI Impact Summit 2026. Mastercard demonstrated India's first authenticated agentic transaction at the summit venue. Pine Labs partnered with OpenAI to announce plans to provide AI agent-driven payment capabilities to over 980,000 merchants. Tewari himself estimated that agentic commerce could add approximately $3 trillion in value to the Indian economy by 2047.

However, concerns about this rapid growth are also emerging. When AI agents make purchasing decisions on behalf of consumers, there is a risk that "why a particular product was chosen" becomes invisible to consumers. Mastercard's analysis identifies three key challenges: the possibility of agents misreading intent and ordering completely different products, the risk of fraud detection systems flagging AI automated transactions as suspicious, and the question of liability when errors occur.

InMobi's Vision: The "Commerce Intelligence Graph"

A notable aspect of Tewari's presentation was the concrete technical architecture being built by Glance, a subsidiary of InMobi.

Tewari explained a structure where multiple AI models collaborate to create agentic experiences. At its core is the "Commerce Intelligence Graph." This is a mechanism similar to a knowledge graph that structures and understands commerce elements such as product information, brands, price ranges, and consumer context. On top of this sits a generative AI experience layer that produces commerce-specialized outputs.

Tewari also introduced the concept of a "Living Commerce Context Graph." This is a mechanism for real-time understanding of consumers' current situation — what they're looking for, their sensitivity to price and brands. "Feed this model with information, and it finds the most efficient purchase path. 'Purchase path optimization' that derives the optimal solution from millions, billions of purchase paths becomes reality," he stated.

InMobi Group has already invested $200 million (approximately ¥30 billion) in Glance AI and its generative AI ad-tech stack. Glance, the consumer-facing business under InMobi, has raised a cumulative $390 million from Google, Jio Platforms, Mithril Capital, and others, reaching a valuation of $1.7 billion. In early U.S. trials, 1.5 million users generated over 40 million personalized looks, with 50% saving or sharing images and 40% initiating the purchase process.

The key point of Tewari's argument is that precisely because of this powerful AI-driven commerce, evolution from "personalized feeds" to "personal feeds" is necessary. Traditional personalized feeds display information selected by corporate algorithms, while personal feeds have AI generate product feeds in real-time for each individual consumer. This distinction is the essence of transparency — the basis for recommendations should be disclosed to consumers, Tewari argues.

Impact on E-Commerce Brands and How to Act

Tewari's arguments and the rapid advancement of agentic commerce offer three important implications for e-commerce brands.

Disclosure of AI agents' "recommendation rationale" is becoming standard. Mastercard's proposed Agentic Token is a mechanism that links AI agents to individual users, ensuring transparency and security throughout the entire transaction process. Mastercard AI Garage's Nitendra Rajput, interviewed by Digit.in, explained that in agentic commerce, "product-finding agents," "user preference-understanding agents," "safe merchant-identifying agents," and "optimal payment method-selecting agents" work collaboratively. E-commerce brands need to organize their product data as structured data that these agents can correctly interpret, and enable traceability of recommendation rationale.

Liability frameworks for errors should be designed in advance. When an AI agent orders a product different from the consumer's intent, who bears the return cost — the retailer, the AI agent developer, or the consumer? Mastercard positions this liability question as the core challenge of agentic commerce and is advancing cross-industry rule development. E-commerce brands should also begin reviewing their return and refund policies for orders placed through AI agents.

India is becoming the testing ground for agentic commerce. India has over 100 million weekly ChatGPT users, making it the world's second-largest market. Pine Labs, Razorpay, and Mastercard are all rapidly advancing agentic transaction pilots, and the best practices established here have the potential to spread across Asia and globally.

Summary

Tewari's argument is significant in that, while being an advocate for agentic commerce, he indicated when to apply the "brakes." In an era where AI takes over commerce decision-making, to avoid repeating the "attention manipulation" mistakes of social media, transparency in recommendation logic, clear liability assignment, and preservation of consumer control are essential.

Key areas to watch include whether Mastercard's Agent Pay and Agentic Token become established as industry standards, what guidelines regulatory authorities in various countries will set for AI agent transactions, and how far InMobi/Glance's Commerce Intelligence Graph will transform actual consumer experiences. Agentic commerce has entered a phase where not only "what to buy" but "how to trust" is being questioned.