Amazon is grappling with what experts are calling the "leader's dilemma" as artificial intelligence reshapes e-commerce. Despite its dominant position in online retail, the company faces a strategic crossroads: whether to resist AI-powered shopping agents that could bypass its platform or embrace them through partnerships to stay ahead. CEO Andy Jassy acknowledged the transformative power of AI earlier this year, telling employees that AI agents would soon influence everyday tasks—from shopping to travel. Four months later, during an earnings call, Jassy confirmed Amazon plans to collaborate with third-party AI agents, engaging in discussions with providers though he did not name them. This shift is underscored by a recent job posting seeking a corporate development leader to drive strategic partnerships in agentic commerce—the emerging trend of AI agents autonomously researching and purchasing products on behalf of users. This move signals Amazon’s recognition that AI is no longer just a tool for internal optimization but a new frontier in customer acquisition and retention. Major tech players have already launched AI shopping tools. OpenAI’s Instant Checkout allows purchases in ChatGPT for select retailers like Walmart, Target, Shopify, and Etsy. Google and Microsoft have introduced similar capabilities, while Perplexity offers an Instant Buy feature in its search engine. These agents aim to streamline shopping by letting users find and buy products without leaving a chat interface. However, the technology remains imperfect. Users report glitches—such as incorrect product images or failed transactions—even when items are in stock. Scot Wingo, founder of e-commerce software firm ReFiBuy, tested Perplexity’s tool and encountered repeated errors while trying to buy a sweater. He also found that ChatGPT once showed a garden rake instead of a coffee machine, highlighting the risks of AI agents pulling in inaccurate or mismatched data. The real concern for Amazon is not just the user experience but the potential loss of control over its most valuable assets. Wingo noted that customer reviews and sales rankings are among Amazon’s most proprietary and powerful data—critical for both consumer trust and AI training. If third-party agents gain access to these insights, they could improve their recommendations while undermining Amazon’s competitive edge. Amazon isn’t standing still. It continues to invest in its own AI tools, particularly Rufus, its in-house chatbot. Since its launch, Rufus has evolved with improved capabilities. The company recently introduced features allowing Rufus to auto-purchase items when prices drop and to suggest products across the web, not just on Amazon. It’s also testing custom shopping guides, mirroring OpenAI’s recent “shopping research” tool. Yet, as retail consultant Jordan Berke points out, Amazon’s size and market dominance make the stakes higher. “Instead of the innovator’s dilemma, I would say Amazon is in what I would call the leader’s dilemma,” he said. “Their market share is so significant that they have the most to lose.” The challenge is balancing openness to AI innovation with the need to protect its core data and ecosystem.