The tech industry is hot for shopping bots, but they're years away

Agentic commerce, the concept of automated software making purchases on behalf of consumers from online merchants, has gained significant attention in the tech and financial industry since early 2025. However, several challenges stand in the way of making agentic commerce a reality. Amazon has blocked Meta's Muse bot and Perplexity's AI bot, citing terms of use restrictions, while Mastercard published a report on the necessary steps to make agentic commerce workable.

The report projects that agentic commerce in US consumer retail could reach $1 trillion in revenue by 2030, and Gartner expects AI agents to handle more than $15 trillion in B2B spending by 2028. Yet, experts remain skeptical about the widespread adoption of agentic commerce within the next five years, citing several factors. First, e-commerce sites already offer automated repeat purchasing, and low-cost items are easy to buy without AI agents.

Second, purchasing high-cost items requires close consideration, which many would be reluctant to delegate to AI agents. Third, trust is crucial in commercial transactions, and AI models have shown a lack of transparency, raising concerns about price manipulation and purchasing decisions. Agentic commerce encompasses more than just the payment moment, including discovery, negotiation, settlement, and post-purchase processes, which makes it even more complex to implement.

To make AI transactions viable, the industry needs deterministic gates to ensure transactions cannot be bypassed by probabilistic models. Additionally, there are practical issues with how e-commerce is currently set up, such as the need for a mechanism to tether a person's identity to an AI agent and addressing the fact that AI agents cannot log into accounts or access necessary data.

Despite these challenges, some experts believe that AI-assisted shopping could become viable within the next five years, but true agentic autonomy remains a distant goal.