Agentic AI is changing the retail landscape by reshaping how consumers shop. Third-party agents, like ChatGPT and Perplexity, are increasingly popular for product research, accounting for up to a quarter of referral traffic for some retailers. However, consumers still prefer retail-owned agents, trusting them three times more. This trust gives retailers a chance to develop on-site AI agents that enhance the shopping experience while also collaborating with third-party agents to stay relevant. The challenge lies in balancing between building their own capabilities and engaging with external platforms, as failure to do so risks disintermediation. Three types of AI agents are emerging in retail: third-party objective agents, proprietary on-site agents, and off-site retailer agents. Amazon's Rufus, for example, is a conversational shopping assistant designed to help customers navigate their product range, while Magalu's Lu operates within WhatsApp to guide users through purchases. Off-site agents, like Amazon's "Buy for Me," allow shoppers to buy from other retailers. The rise of these agents poses both a threat and an opportunity for traditional retailers. Those who can leverage their unique data and expertise to create compelling on-site agents will be better positioned to attract and retain customers. Retailers need to act swiftly. They must build capabilities where they have a unique advantage and consider strategic partnerships with larger agents. Collaborations, such as those initiated by Walmart and Shopify with OpenAI, can help retailers maintain control over customer data and optimize their visibility in a crowded market. As generative AI becomes the default starting point for many consumers, retailers must adapt their strategies to ensure they remain relevant in this evolving environment.