New York, 29/09/2025: Imagine you need groceries, or want to buy a birthday gift, or a new pair of running shoes, or shades to get eyeballs at a tony party, only that you find it a bit bothersome to go to an online shopping website and make a specific choice, checking testimonials and price tags.

What if someone is there to do all these for you while you are as snug as a bug in a rug! Sounds crazy, doesn't it? Time to get out of the slumber, for, in not so distant future, an AI agent will do the whole shopping for you —browsing products, comparing prices, and making the purchase without you even lifting a finger. This vision of agentic commerce, the biggest change in retail since e-commerce itself, is more up close and personal than you would have ever thought.

Unlike today's chatbots that just respond to customer queries, agentic AI is proactive, making decisions and solving complex problems without human intervention. These digital agents can understand your preferences, anticipate your needs, and execute complex shopping tasks from start to finish without constant human guidance.

The government is recognizing the impact of AI on commerce and the broader economy. While announcing changes to the U.S. AI Safety Institute, Commerce Secretary Howard Lutnick's statement that "AI will bring transformative advances that will boost U.S. economic and national security" makes a prescient point.

Leading the pack of technocrats on this road to massive e-commerce transformation is Sudarshan Prasad in Nagavalli, Lead Product Manager at PayPal and an independent AI researcher whose work in predictive analytics and federated learning has predicted many of the innovations we are seeing in agentic commerce. In his research, 'Predictive Analytics in eCommerce: AI-Driven Insights for Market Trends and Consumer Behavior,' Sudarshan has outlined how AI-powered systems could change online shopping through sophisticated pattern recognition and personalization engines.

Sudarshan Prasad Nagavalli

"The digitalization that has happened today has made it easier to identify markets and how customers behave," Sudarshan says in his research. His work predicts exactly what we're seeing today—AI systems that can process vast amounts of transactional data, social media patterns, and browsing behavior to make highly accurate purchasing decisions on behalf of consumers.

The system works through advanced authorization protocols. When a customer approves a purchase—say, a $1,000 laptop—they're essentially providing a "Purchase Signal" that authorizes their AI agent to complete that specific transaction. The AI Agent will then autonomously work through the checkout process and, before completing the transaction, compare the price to the original authorization. If it's a match, the agent can autonomously complete the transaction.

Sudarshan's research into explainable AI is particularly relevant here. His work on transparency in AI decision-making addresses a critical question in agentic commerce: how can consumers trust an AI system to make purchasing decisions on their behalf? His research shows how AI systems can provide clear explanations for their recommendations and actions, building the foundation of trust necessary for autonomous commerce to succeed.

Big tech and financial companies are building the infrastructure for agentic commerce. Google, PayPal, Perplexity, Visa, and Mastercard are leading the Agentic AI payments charge.

Google has launched AI Shopping Mode, which allows users to browse 50 billion product listings and check out autonomously. Visa has introduced Intelligent Commerce, which enables AI to deliver personalized shopping at scale. Mastercard has followed with Agent Pay, which embeds purchase into AI-generated product recommendations. Mastercard is pioneering a new payment model where AI, not the consumer, initiates the purchase.

Researchers and industry experts identify two types of AI-powered shopping experiences. On the utility side, we'll have agents that do the dirty work. They hunt for good deals. They re-order our toilet paper. They know mom's birthday is coming up, and they're ready to go with a shortlist of gift ideas.

This utility shopping—mundane, repetitive purchases that require significant cognitive load—represents the initial frontier for agentic commerce. The emotional side of shopping, in my view, will be less agent-centric. Here, people like shopping, and they don't want to delegate. Instead, LLMs can better recommend products. AI = personalization. Sudarshan's research is relevant to both.

McKinsey & Company's latest research is a reality check alongside the huge potential. Nearly 8 in 10 companies are using gen AI—yet just as many are seeing no significant bottom-line impact. This "gen AI paradox" is due to an imbalance between widely deployed but low-impact applications and vertical use cases stuck in the pilot phase.

But McKinsey estimates gen AI could yield $4.4 trillion in productivity growth over the long term. They see agentic AI as the solution to this paradox. By automating complex business workflows, agents unlock the full potential of vertical use cases. Forward-thinking companies are already using agents to transform core processes.

Despite the promise, there are still many obstacles. While eCommerce companies will start testing AI agents in 2025, the technology won't revolutionize the industry for several more years, according to a senior analytics executive at a fulfillment company. Challenges include scalability issues, complex data integration requirements, and fundamental questions around consumer trust in AI decision-making.

Payment infrastructure is another big bottleneck. Until recently, this model of commerce was limited by payment functionality. AI Agents could complete most of the shopping journey, including browsing for products and services and returning suggestions based on the customer-provided prompt, up until the payment itself.

Security and fraud prevention add another layer of complexity. Sudarshan's expertise in federated learning and zero-trust security architectures becomes particularly relevant here. His research shows how AI systems can be secure while operating autonomously, providing the foundation for secure agentic commerce transactions.

According to Gartner, 33% of enterprises will have agentic AI by 2028, up from less than 1% today.

The transformation will follow a predictable pattern. Simple, repetitive purchases will be automated first—subscription renewals, basic household items, and routine business supplies. More complex, higher-stakes purchases will remain human-directed but with significant AI assistance.

Sudarshan's research provides a comprehensive blueprint for agentic commerce systems. His framework has four foundational pillars for successful autonomous shopping platforms:

1. Intelligent Threat Detection and Response: Drawing from Sundar Tiwari's cybersecurity expertise, Sudarshan's research recommends real-time monitoring systems that can detect and neutralize fraudulent transactions before they compromise the autonomous shopping experience.
2. Privacy-Preserving Data Collaboration: His federated learning approach enables AI agents to learn from distributed data sources across multiple retailers while keeping strict privacy boundaries—essential for building consumer trust in autonomous purchasing decisions.
3. Adaptive Decision Architecture: Sudarshan's work shows how AI systems must evolve their decision-making processes based on changing market conditions, consumer preferences, and emerging shopping patterns.
4. Transparent Authority Validation: His explainable AI research ensures every autonomous purchase decision can be traced, understood, and validated—giving consumers clear explanations for why their AI agent made specific buying decisions.

"Organizations must move from current reactive security practices to proactive defense mechanisms with better threat detection and response capabilities," Sudarshan's research concluded, a principle now fundamental to agentic commerce implementation.

The ultimate vision goes beyond human-to-agent interactions to agent-to-agent commerce. The next phase of agentic commerce may see AI Agents accepting and processing payments on behalf of merchants. In this scenario, a consumer's AI agent could directly negotiate with a retailer's AI agent, potentially leading to dynamic pricing, personalized promotions, and entirely new forms of commerce.

As agentic commerce moves from experimental to mainstream, the question isn't if AI agents will change shopping, but how fast and how much. Sudarshan's research provides the theoretical foundation and practical frameworks to navigate this transition.

The companies and consumers who understand and adapt to this agentic future will be at the forefront of the biggest commercial revolution since the internet itself. As we stand at the threshold of this transition, one thing is clear: the future of shopping will be intelligent, autonomous, and fundamentally different from anything we've seen before.